WO2024217301A1 - Point cloud coding processing method, point cloud decoding processing method and related device - Google Patents

Abstract

The present application belongs to the technical field of computers. Disclosed are a point cloud coding processing method, a point cloud decoding processing method and a related device. The point cloud coding processing method comprises: determining a target distance between a target layer and a root node of a transform tree corresponding to a point cloud to be coded; determining a target prediction residual value on the basis of the target distance; and coding the target prediction residual value, wherein when the target distance is less than or equal to a first preset distance, the target prediction residual value comprises a first prediction residual value obtained by performing inter-frame prediction on attribute information of nodes of the target layer; or when the target distance is greater than the first preset distance and less than or equal to a second preset distance, the target prediction residual value comprises a second prediction residual value obtained by performing prediction processing on the attribute information of the nodes of the target layer; or when the target distance is greater than the second preset distance, the target prediction residual value comprises a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the nodes of the target layer.

Description

Point cloud encoding processing method, point cloud decoding processing method and related equipment

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims priority to Chinese Patent Application No. 202310408599.1 filed in China on April 17, 2023, the entire contents of which are incorporated herein by reference.

Technical Field

The present application belongs to the field of computer technology, and specifically relates to a point cloud encoding processing method, a point cloud decoding processing method and related equipment.

Background Art

Point cloud is a representation of a three-dimensional object or scene. It is composed of a set of discrete points that are irregularly distributed in space and express the spatial structure and surface properties of the three-dimensional object or scene. In order to accurately reflect the information in space, the number of discrete points required is quite large. In order to reduce the bandwidth occupied by point cloud data storage and transmission, the point cloud data needs to be encoded and compressed. Point cloud data usually consists of geometric information describing the location, such as three-dimensional coordinates (x, y, z) and attribute information of the location, such as color (R, G, B) or reflectivity. In the process of point cloud coding and compression, the encoding of geometric information and attribute information is performed separately.

In the related technology, in the process of encoding attribute information based on the Region Adaptive Hierarchal Transform (RAHT) algorithm, the tree structure corresponding to the point cloud to be encoded is divided into two parts. The nodes in the upper layer are predicted using the inter-frame prediction algorithm, and the nodes in the lower layer are predicted using the intra-frame prediction algorithm. This has poor flexibility and leads to low coding efficiency.

Summary of the invention

The embodiments of the present application provide a point cloud encoding processing method, a point cloud decoding processing method and related equipment, which can solve the problem of low encoding efficiency.

In a first aspect, a point cloud encoding processing method is provided, which is executed by an encoding end and includes:

Determine the target distance between the target layer and the root node of the transform tree corresponding to the point cloud to be encoded;

Determining a target prediction residual value based on the target distance;

Encoding the target prediction residual value to obtain a first encoding result;

The code stream of the point cloud to be encoded includes the first encoding result;

In a case where the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or

In a case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or

When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer.

In a second aspect, a point cloud decoding processing method is provided, which is executed by a decoding end and includes:

Determine the target distance between the target layer and the root node of the transform tree corresponding to the point cloud to be decoded;

Decode the first encoding result in the bitstream of the point cloud to be decoded to obtain a target prediction residual value;

Acquire attribute information of nodes in the target layer based on the target prediction residual value and the target distance;

Wherein, when the target distance is less than or equal to the first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or

In a case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or

When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer.

In a third aspect, a point cloud coding processing device is provided, comprising:

A first determination module is used to determine a target distance between a target layer and a root node of a transform tree corresponding to a point cloud to be encoded;

A second determination module, used to determine a target prediction residual value based on the target distance;

An encoding module, used for encoding the target prediction residual value to obtain a first encoding result;

The code stream of the point cloud to be encoded includes the first encoding result;

In a case where the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or

In a case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or

When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer.

In a fourth aspect, a point cloud decoding processing device is provided, comprising:

A determination module, used to determine a target distance between a target layer and a root node of a transform tree corresponding to a point cloud to be decoded;

A decoding module, used for decoding a first encoding result in a bit stream of a point cloud to be decoded to obtain a target prediction residual value;

An acquisition module, used for acquiring attribute information of the nodes of the target layer based on the target prediction residual value and the target distance;

Wherein, when the target distance is less than or equal to the first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or

In the case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or

When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer.

In a fifth aspect, a terminal is provided, which includes a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the method described in the first aspect are implemented, or the steps of the method described in the second aspect are implemented.

In a sixth aspect, a terminal is provided, comprising a processor and a communication interface, wherein the processor is used to: determine a target distance between a target layer and a root node of a transform tree corresponding to a point cloud to be encoded; determine a target prediction residual value based on the target distance; encode the target prediction residual value to obtain a first encoding result; wherein the code stream of the point cloud to be encoded includes the first encoding result; when the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by inter-frame prediction of attribute information of a node of the target layer; or when the target distance is greater than the first preset distance and less than or equal to a second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer; or when the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by intra-frame prediction of the attribute information of the node of the target layer.

In the seventh aspect, a terminal is provided, comprising a processor and a communication interface, wherein the processor is used to: determine a target distance between a target layer and a root node of a transform tree corresponding to a point cloud to be decoded; decode a first encoding result in a code stream of the point cloud to be decoded to obtain a target prediction residual value; obtain attribute information of the nodes of the target layer based on the target prediction residual value and the target distance; wherein, when the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by inter-frame prediction of the attribute information of the nodes of the target layer; or when the target distance is greater than the first preset distance and less than or equal to a second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the nodes of the target layer; or when the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by intra-frame prediction of the attribute information of the nodes of the target layer.

In an eighth aspect, a readable storage medium is provided, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented, or the steps of the method described in the second aspect are implemented.

In the ninth aspect, a coding and decoding system is provided, comprising: an encoding end device and a decoding end device, wherein the encoding end device can be used to execute the steps of the method described in the first aspect, and the decoding end device can be used to execute the steps of the method described in the second aspect.

In the tenth aspect, a chip is provided, comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the method described in the first aspect, or to implement the method described in the second aspect.

In an eleventh aspect, a computer program/program product is provided, wherein the computer program/program product is stored in a storage medium, and the program/program product is executed by at least one processor to implement the steps of the method according to the first aspect, Or implement the steps of the method as described in the second aspect.

In an embodiment of the present application, a target prediction residual value is determined based on the target distance, and the target prediction residual value is encoded. When the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by inter-frame prediction of the attribute information of the node of the target layer; or when the target distance is greater than the first preset distance and less than or equal to a second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer; or when the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by intra-frame prediction of the attribute information of the node of the target layer. By comparing the target distance with the first preset distance and the second preset distance, different prediction methods are determined, which has high flexibility and can improve coding efficiency.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a flow chart of a prediction method in the related art;

FIG2 is a schematic diagram of a node in the related art;

FIG3 is a schematic diagram of binary RAHT decomposition of a node block in the related art;

FIG4 is a schematic diagram of a prediction method in the related art;

FIG5 is a flowchart of a point cloud coding processing method provided in an embodiment of the present application;

FIG6 is a flowchart of a point cloud decoding method provided in an embodiment of the present application;

FIG. 7 is a second flowchart of a point cloud coding processing method provided in an embodiment of the present application;

FIG8 is a second flowchart of a point cloud decoding processing method provided in an embodiment of the present application;

FIG9 is a schematic diagram of the structure of a point cloud coding processing device provided in an embodiment of the present application;

FIG10 is a schematic diagram of the structure of a point cloud decoding processing device provided in an embodiment of the present application;

FIG11 is a schematic diagram of the structure of a communication device provided in an embodiment of the present application;

FIG. 12 is a schematic diagram of the structure of a terminal provided in an embodiment of the present application.

DETAILED DESCRIPTION

The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of this application.

The terms "first", "second", etc. of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the terms used in this way are interchangeable where appropriate, so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of one type, and the number of objects is not limited, for example, the first object can be one or more. In addition, "or" in the present application represents at least one of the connected objects. For example, "A or B" covers three schemes, namely, Scheme 1: including A but not including B; Scheme 2: including B but not including A; Scheme 3: including both A and B. The character "/" generally indicates that the objects associated with each other are in an "or" relationship.

The term "indication" in this application can be a direct indication (or explicit indication) or an indirect indication (or implicit indication). A direct indication can be understood as the sender explicitly informing the receiver of specific information, operations to be performed, or request results in the sent indication; an indirect indication can be understood as the receiver determining the corresponding information according to the indication sent by the sender, or making a judgment and determining the operation to be performed or the request result according to the judgment result.

The codec end corresponding to the point cloud codec processing method in the embodiment of the present application may be a terminal, which may also be referred to as a terminal device or a user terminal (User Equipment, UE). The terminal may be a mobile phone, a tablet computer (Tablet Personal Computer), a laptop computer (Laptop Computer) or a notebook computer, a personal digital assistant (Personal Digital Assistant, PDA), a handheld computer, a netbook, an ultra-mobile personal computer (ultra-mobile personal computer, UMPC), a mobile Internet device (Mobile Internet Device, MID), an augmented reality (augmented reality, AR)/virtual reality (virtual reality, VR) device, a robot, a wearable device (Wearable Device) or a vehicle-mounted device (Vehicle User Equipment, VUE), a pedestrian terminal (Pedestrian User Equipment, PUE) and other terminal side devices, and the wearable device includes: a smart watch, a bracelet, a headset, glasses, etc. It should be noted that the specific type of the terminal is not limited in the embodiment of the present application.

For ease of understanding, some contents involved in the embodiments of the present application are described below:

1. Geometry-based Point Cloud Compression (GPCC) encoding and decoding

For the GPCC encoded and decoded point cloud, its geometric information and attribute information are encoded separately. First, the geometric information is encoded, and then the attribute information is encoded using the reconstructed geometric information.

There are two main ways to encode attribute information, namely, level of detail-based prediction and lifting transform and Region Adaptive Hierarchal Transform (RAHT).

The attribute coding method of RAHT can be divided into intra-frame RAHT and inter-frame RAHT according to the prediction method adopted.

Intra-frame prediction uses only the nodes that have been encoded in the current frame to predict the current node, while inter-frame prediction uses the nodes of the reference frame to predict the current node.

a) Intra-frame RAHT technology:

Intra-frame RAHT technology is called regional adaptive transformation based on upsampling prediction, which is transformed layer by layer from top to bottom. In each layer, RAHT is applied to each unit node containing 2×2×2 child nodes. When encoding the unit node of the current layer with 8 child nodes, its parent node (previous layer), coplanar neighbor parent node (previous layer), collinear neighbor parent node (previous layer), and encoded coplanar neighbor child node (same layer) and encoded collinear neighbor child node (same layer) are used to perform weighted prediction on the attributes of the placeholder child node of the node. After that, the obtained attribute prediction value and the true value are RAHT transformed and the difference is made to obtain the AC coefficient (AC coefficient) residual, and the AC coefficient residual is quantized and encoded. However, when the number of child nodes to be encoded, the number of neighbor grandparent nodes and the number of neighbor parent nodes do not meet certain conditions, intra-frame prediction will not be enabled, and the original attribute value of the current node to be encoded will be directly RAHT transformed, and then the transform coefficient will be quantized and encoded.

The following is the specific process of intra-frame RAHT technology:

The first step is to build a transformation tree structure. Starting from the bottom layer, the octree structure is built from the bottom up. In the process of building the transformation tree, it is necessary to generate corresponding Morton code information, attribute information and weight information for the merged nodes.

The second step is to parse the tree from top to bottom. Starting from the root node, upsampling prediction and regional adaptive hierarchical transform (RAHT) are performed layer by layer.

If the current node is a root node, no upsampling prediction is performed, and the attribute information of the node is directly subjected to RAHT transformation to obtain the DC coefficient and AC coefficient.

If it is not the root node, it is necessary to determine whether to predict the current eight child nodes, as shown in Figure 1:

1. Determine whether the number of placeholder child nodes NumValidc is equal to 1: If the number of placeholder child nodes (non-empty) of the current node to be encoded is 1, set the number of its neighbor parent nodes NumValidP to val (for example, 19), do not make predictions, directly perform RAHT transformation on the original attribute information of the current node, and then quantize and entropy encode the obtained AC coefficients.

Otherwise continue with step 2:

2. Determine whether the number of neighboring grandparent nodes NumValidGP is greater than or equal to TH1 (for example, 2): If the number of neighboring grandparent nodes (including grandparent nodes) of the current subnode to be encoded is less than 2, no prediction is performed, and the original attribute information of the current node is directly RAHT transformed, and then the obtained AC coefficients are quantized and entropy encoded. Otherwise, proceed to step 3:

3. Neighbor search: The search range is: the parent node of the current child node to be encoded (1), the coplanar neighbor node of the parent node of the current child node to be encoded (6), the colinear neighbor node of the parent node of the current child node to be encoded (12), the coplanar neighbor node of the current child node to be encoded (6), the colinear neighbor node of the current child node to be encoded (12). Search the above neighbor nodes in turn. If the neighbor node exists, record its corresponding index information and the number of neighbors of the parent node (including the parent node itself). Then proceed to step 4.

4. Determine whether the number of neighboring parent nodes is greater than or equal to TH2 (for example, 6): If the number of neighboring parent nodes of the current child node to be encoded is less than TH2, no weighted prediction is performed, and the original attribute information of the current node to be encoded is directly RAHT transformed, and then the obtained AC coefficients are quantized and entropy encoded. Otherwise, weighted prediction is performed.

The third step is weighted prediction. The nearest neighbors found in the neighbor search are used to perform weighted prediction on each child node of the current node to be encoded, as shown in Figure 2.

The prediction weight of the parent node is specified to be 9, the prediction weight of the neighboring child node coplanar with the current child node to be encoded is 5, the prediction weight of the neighboring child node colinear with the current child node to be encoded is 2, the prediction weight of the neighboring parent node coplanar with the current child node to be encoded is 3, and the prediction weight of the neighboring parent node colinear with the current child node to be encoded is 1.

Among them, the parent node can be directly used to predict each child node of the current block to be encoded. For other neighbor nodes (neighbor parent nodes and encoded neighbor child nodes), it is necessary to perform a step to determine whether they can be used to predict the child nodes of the current block. The determination steps are as follows:

1. As shown in Figure 2, for neighbor parent nodes with indexes 1-7, since their child nodes have not been encoded yet, only the neighbor parent nodes can be used for prediction.

First, determine whether the neighbor parent node of the position exists. If not, continue to determine the next neighbor parent node.

If it exists, it is necessary to determine whether the current neighbor parent node can be used for prediction and to screen the neighbor parent node. The whole process is as follows:

1) According to the attribute value of the parent node, two prediction thresholds are set to further filter the nearest neighbors and filter out unreasonable neighbor parent nodes to improve the accuracy of the prediction. Let the two thresholds be limitLow and limitHigh, and let the attribute value of the parent node be attrPar, then:
limitLow＝attrpar*2
limitHigh=attrpar*25

Assume that the attribute value of the current neighbor's parent node is attrNei, and make the following judgments on it:
limitLow<attrnei*10<limitHigh

If the condition is not met, the current neighbor parent node cannot be used to predict the child node of the current block to be encoded; if the condition is met, continue to make the following judgment;

2) Determine whether the current neighboring parent node meets the condition of being coplanar and colinear with each of the current child nodes to be encoded. If this condition is not met, the current neighboring node cannot be used to perform weighted prediction on the current child node to be encoded; if this condition is met, the current neighboring node is used to perform weighted prediction on the current child node to be encoded.

2. For neighbor parent nodes with indexes 8-19, since their child nodes have been encoded, if there is a neighbor child node at the same position, it can be used to replace the neighbor parent node to which it belongs for prediction, which will have a better prediction effect.

First, determine whether the neighbor parent node of the position exists. If not, continue to determine the next neighbor parent node.

If it exists, it is necessary to determine whether the current neighbor parent node can be used for prediction and screen the neighbor parent node. The specific process is as follows:

1) According to the attribute value of the parent node, two prediction thresholds are set to further filter the nearest neighbors and filter out unreasonable neighbor parent nodes to improve the accuracy of the prediction. Suppose the attribute value of the current neighbor parent node is attrNei, and make the following judgment on it:
limitLow<attrnei*10<limitHigh

If the condition is not met, the current neighbor parent node cannot be used to predict the child node of the current block to be encoded; if the condition is met, the following judgment is continued:

2) Determine whether there is a neighbor child node in the same layer as the current parent node's neighbor: For coplanar neighbor parent nodes, if one of the child nodes is also coplanar with the current child node to be predicted, use the coplanar neighbor child node to replace the coplanar neighbor parent node to make a weighted prediction for the child node to be predicted; for collinear neighbor parent nodes, if one of the child nodes is also collinear with the current predicted child node, use the collinear neighbor child node to replace the collinear neighbor parent node for weighted prediction. If there is no such child node, continue to use the neighbor parent node for weighted prediction.

3. Finally, each child node of the current node to be encoded uses the neighboring nodes that meet the conditions as the reference point set to perform weighted prediction and obtain the attribute prediction value. Then, the original attribute value and the predicted attribute value are RAHT transformed and the difference is obtained to obtain the AC coefficient residual, which is quantized and entropy coded.
Attr _res =AttrCoff _org -AttrCoff _pred

The fourth step is RAHT transformation. The process is as follows:

Perform RAHT transformation on the original attribute value or (and) attribute prediction value; if the current block is not predicted, only The original attribute value is subjected to RAHT; if the current block is predicted, both the original attribute value and the attribute prediction value need to be subjected to RAHT transformation.

When performing RAHT transformation, the original attribute values and/or attribute prediction values need to be normalized first, and then the processed values are subjected to RAHT transformation.

First, normalize the original attribute value. Suppose the original attribute value of the current child node is A _i and the weight is w _i (the size is the number of points contained in the current node), then:

If prediction is used, let the attribute prediction value of the current child node be A _pi , then:

Perform RAHT transformation on the processed attribute prediction value A′ _pi or the original attribute value A′ _i. For a 2*2*2 node block, transform it along three directions respectively, and perform four transformations in each direction. Therefore, each 2*2*2 block performs the following transformations in total:

1) Transform along the first direction to obtain low-frequency L nodes and high-frequency H nodes;

2) Transform the L nodes and H nodes along the second direction to obtain low-frequency LL nodes and high-frequency LH, HL, and HH nodes;

3) Transform the LL, LH, HL, and HH nodes along the third direction to obtain low-frequency LLL nodes and high-frequency LLH, LHL, LHH, HLL, HLH, HHL, and HHH nodes.

Among them, LLL is the DC coefficient, and LLH, LHL, LHH, HLL, HLH, HHL, and HHH are AC coefficients.

The AC coefficients are quantized and entropy coded in the order of LLH, LHL, HLL, LHH, HLH, HHL, HHH.

Due to the sparsity of point clouds, in general, each 2*2*2 node block usually has less than 8 occupied child nodes, so not all AC coefficients will exist, and non-existent AC coefficients will not be encoded.

The specific process is shown in Figure 3:

Specifically, when performing two-point transformation, assuming that the input attribute values are T ₀₁ and T ₁₁ , respectively, and the transformation coefficients are T ₁ and T _w0+1 , the transformation formula used is:

Where a and b are calculated by the weight of the current node, and the weight of the current node is obtained from the weight array:

If the current node is the root node, no prediction is performed, and the DC coefficient and AC coefficient of the original attribute value after transformation are quantized and entropy coded;

If the current node is not a root node and no prediction is performed, the AC coefficients of the original attribute values after transformation are quantized and entropy encoded;

If the current node is not a root node and prediction is performed, the residual of the AC coefficient of the original attribute value after transformation and the AC coefficient of the attribute prediction value after transformation is calculated, and the residual is quantized and entropy encoded.

At the decoding end, the coefficients are decoded and dequantized, and then the RAHT inverse transform is performed.

The inverse RAHT transform is the inverse process of the RAHT transform. The inverse transform formula is as follows:

Wherein, T ₁ and T _w0+1 are transformation coefficients, T ₀₁ and T ₁₁ are reconstruction attribute values, and the calculation method of a and b is the same as the calculation method of a and b in the transformation process.

If the current node is the root node, then directly perform the inverse transformation to obtain the attribute reconstruction value of each child node;

If the current node is not a root node and no prediction is performed, the DC coefficient inherits the attribute reconstruction value of the parent node, and then performs an inverse transformation to obtain the attribute reconstruction value of each child node;

If the current child node is the root node and prediction is performed, the DC coefficient inherits the attribute reconstruction value of the parent node, and the AC coefficient is the AC coefficient residual plus the AC coefficient of the attribute prediction value. Then an inverse transformation is performed to obtain the attribute reconstruction value of each child node.

b) Inter-frame RAHT technology:

The inter-frame RAHT technology uses the inter-frame prediction attribute values of some qualified nodes in the reference frame to predict the attributes of the nodes in the current frame, and then performs RAHT transformation on the attribute prediction value and the true attribute value respectively and makes a difference, and quantizes and encodes the AC coefficient residual.

The following is a detailed introduction to the inter-frame scheme:

Inter-RAHT is a method for inter-frame prediction of DC and AC coefficients between RAHT frames: inter-frame prediction is only used for the first five layers of nodes, and the remaining layer nodes only use intra-frame prediction methods.

In the first five layers (high-level nodes):

For the DC coefficient:

The DC coefficient prediction residual DC _residual is: the DC coefficient DC _current of the root node of the current frame minus the DC coefficient DC _reference of the root node of the reference frame.
DC _residual =DC _current -DC _reference

For AC coefficients:

The prediction process is as follows:

1. Reference frame reconstruction: Merge the points in the reference frame within a certain range of the current point into one point, add the attributes, and take the coordinates of the current point. Then, the same prediction tree construction and analysis can be performed. The specific process is as follows:

a) For each point, there is a reconstructed reference point with the same position coordinates. The attributes of this reference point are the sum of the attributes of the points within a certain range in the reference frame obtained according to certain rules.

b) The rule is: if the distance of a point in the reference frame to the current point of the current frame is closer than the Morton code distance to the next point of the current frame, then this point is determined to be used to reconstruct the current point, the attribute value is accumulated, and the weight is increased by one. Until the condition is not met, continue to determine the next reconstruction point.

c) In addition, the start and end distances are specified. If the distance is too far, it will be considered that the attribute values are too different and not suitable for reconstruction.

Only points in the reference frame whose distance to the first point of the current frame is no greater than 64 will be used for reconstruction.

Only points in the reference frame whose distance to the last point of the current frame is no greater than 64 will be used for reconstruction.

2. Tree construction: The reconstructed reference frame and the current frame are subjected to the bottom-up operation of constructing the prediction tree, and the same results are obtained. The prediction tree is constructed to generate the attribute value of the current frame node and the predicted attribute value and weight value of the corresponding reference frame node.

3. As shown in Figure 4, the predicted attribute value is selected: in the first five layers, if the attribute value of the node at the corresponding position of the reference frame is not 0 and is within 20% to 250% of the attribute value of the current parent node, the attribute value Attr _{predicted_inter} of the node in the reference frame is used as the attribute prediction value, otherwise the intra-frame prediction value Attr _{predicted_intra} of the current frame is used as the attribute prediction value. Perform RAHT transformation on the predicted attribute value Attr _predicted to obtain the AC coefficient of the predicted value. Subtract the AC coefficient AC _current obtained by the current frame transformation from the predicted AC coefficient AC _predicted to obtain the AC coefficient prediction residual AC _residual , and quantize and encode the residual. The calculation formula is as follows:
AC _residual =AC _current -AC _predicted
AC _predicted =Transform(Attr _predicted )
Attr _predicted =Attr _{predicted_inter} ? Attr _{predicted_inter} :Attr _{predicted_intra}

4. The remaining layers only use the original intra-frame RAHT prediction method.

In the related technology, in the process of encoding attribute information based on the region adaptive hierarchical transform (RAHT) algorithm, the tree structure corresponding to the point cloud to be encoded is divided into two parts. The nodes in the upper layer are predicted using the inter-frame prediction algorithm, and the nodes in the lower layer are predicted using the intra-frame prediction algorithm. The flexibility is poor, resulting in low coding efficiency.

In the following, in combination with the accompanying drawings, the point cloud encoding processing method, the point cloud decoding processing method and related equipment provided in the embodiments of the present application are described in detail through some embodiments and their application scenarios.

Referring to FIG. 5 , FIG. 5 is a flow chart of a point cloud coding processing method provided in an embodiment of the present application, which can be applied to the coding end. As shown in FIG. 5 , the point cloud coding processing method includes the following steps:

Step 101: Determine a target distance between a target layer and a root node of a transform tree corresponding to a point cloud to be encoded.

Among them, the geometric information and attribute information of the point cloud to be encoded can be obtained; based on the geometric information of the point cloud to be encoded, a transform tree corresponding to the point cloud to be encoded is generated. The transform tree corresponding to the point cloud to be encoded can include multiple layers, the root node can be considered as the first layer of the transform tree, and the target layer can be any layer in the transform tree. The target distance can refer to the number of layers from the target layer to the root node. For example, the target distance between the second layer of the transform tree and the root node can be one layer, the target distance between the third layer of the transform tree and the root node can be two layers, the target distance between the fourth layer of the transform tree and the root node can be three layers, and so on.

Step 102: determining a target prediction residual value based on the target distance;

Step 103: Encode the target prediction residual value to obtain a first encoding result;

The code stream of the point cloud to be encoded includes the first encoding result;

In a case where the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or

In a case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or

When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer.

The first preset distance and the second preset distance may both be the number of layers from the root node. The first preset distance is smaller than the second preset distance. For example, the first preset distance may be 5 layers and the second preset distance may be 10 layers; or the first preset distance may be 7 layers and the second preset distance may be 13 layers; or the first preset distance may be 8 layers and the second preset distance may be 15 layers; etc. This embodiment does not limit the first preset distance and the second preset distance.

It should be noted that, when the target distance is less than or equal to the first preset distance, the node of the target layer is the node of the upper layer that is closer to the root node. Since the nodes of the upper layer of the transform tree are larger and contain more points, even if there is movement between the two frames, there will basically not be much change in these upper layer nodes. At this time, if inter-frame prediction is used, the inter-frame attribute prediction residual may be smaller than the intra-frame attribute prediction residual, so that the use of inter-frame prediction in the upper layer will have a better effect.

In addition, when the target distance is greater than the first preset distance and less than or equal to the second preset distance, the inter-frame prediction method or the intra-frame prediction method can be determined by a rate-distortion optimization algorithm.

In addition, when the target distance is greater than the second preset distance, the node of the target layer is a node of the lower layer that is farther away from the root node. Since the lower layer block is small and contains fewer points, when there is a certain amount of motion between the two frames, the number of points contained in each block may change greatly, which is not suitable for inter-frame prediction. Therefore, using intra-frame prediction in the lower layer will have a better effect.

It should be noted that the inter-frame prediction method in the related art is to determine whether the current layer has turned on the inter-frame prediction mode by the distance from the root node (number of layers). That is, when the distance from the current layer to the root node is less than or equal to the set threshold, the corresponding inter-frame prediction algorithm is applied. When the distance from the current layer to the root node is greater than the set threshold, only the existing intra-frame prediction method is used. However, for some sequences, the results of inter-frame prediction for some nodes greater than the threshold may have smaller residuals than intra-frame prediction, and the prediction effect is better.

This embodiment adopts a more flexible method to select the method of using the intra-frame and inter-frame prediction modes, instead of simply using the number of layers from the root node as the basis, the divided octree is divided into two, the inter-frame prediction is used for the nodes in the upper layer, and the intra-frame prediction is generally used for the remaining nodes. This embodiment uses a more flexible method to determine the prediction mode used by the current node, so as to make the prediction residual smaller.

In an embodiment of the present application, a target prediction residual value is determined based on the target distance, and the target prediction residual value is encoded. When the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by inter-frame prediction of the attribute information of the node of the target layer; or when the target distance is greater than the first preset distance and less than or equal to a second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer; or when the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by intra-frame prediction of the attribute information of the node of the target layer. By comparing the target distance with the first preset distance and the second preset distance, different prediction methods are determined, which has high flexibility and can improve coding efficiency.

Optionally, when the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer based on a rate-distortion optimization algorithm.

Wherein, when the target distance is greater than the first preset distance and less than or equal to the second preset distance, A first generation value corresponding to the inter-frame residual coefficient and a second generation value corresponding to the intra-frame residual coefficient may be determined based on a rate-distortion optimization algorithm, and a second prediction residual value may be determined based on the first generation value and the second generation value.

In this implementation, when the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer based on a rate-distortion optimization algorithm. In this way, it supports judging the cost of using intra-frame prediction and inter-frame prediction through a rate-distortion optimization algorithm, thereby selecting a suitable prediction method for prediction processing to obtain a second prediction residual value.

Optionally, determining a target prediction residual value based on the target distance includes:

When the target distance is greater than the first preset distance and less than or equal to the second preset distance, determine the inter-frame residual coefficient and the intra-frame residual coefficient of the attribute information of the node of the target layer;

Determine a first generation value corresponding to the inter-frame residual coefficient and a second generation value corresponding to the intra-frame residual coefficient based on a rate-distortion optimization algorithm;

determining a second prediction residual value based on the first generation value and the second generation value;

Among them, the target prediction residual value includes the second prediction residual value.

Here, cost can also be described as expenditure.

In one implementation, the second generation value Cost_intra corresponding to the intra-frame residual coefficient and the first generation value Cost_inter corresponding to the inter-frame residual coefficient may be calculated respectively, and the second generation value of the intra-frame residual coefficient or the first generation value of the inter-frame residual coefficient may be obtained by the following formula:
Cost＝D+λ*R

Wherein, D (distortion) is the sum of absolute differences (Sum of Absolute Difference, SAD) between the reconstructed attribute and the original attribute obtained after using the prediction residual coefficient (intra-frame residual coefficient or inter-frame residual coefficient).

Here, R (rate: bit rate) is the number of bits required to encode the prediction residual coefficient (intra-frame residual coefficient or inter-frame residual coefficient).

The value of λ depends on the quantization parameter QP, which is calculated as follows:

It should be noted that the process of determining the inter-frame residual coefficients of the attribute information of the nodes of the target layer may include: determining the inter-frame prediction value of the attribute information of the nodes of the target layer, transforming the inter-frame prediction value to obtain the inter-frame transformation coefficient, and obtaining the inter-frame residual coefficient based on the inter-frame transformation coefficient; or, may include: if there is a transformation coefficient of a reference node having the same position as the node of the target layer in the reference frame of the point cloud to be encoded, then determining the transformation coefficient of the reference node having the same position as the inter-frame transformation coefficient of the attribute information of the node, and determining the inter-frame residual coefficient of the attribute information of the node based on the transformation coefficient of the node and the inter-frame transformation coefficient of the attribute information of the node.

In addition, determining the intra-frame residual coefficient of the attribute information of the node of the target layer may include: determining the intra-frame prediction value of the attribute information of the node of the target layer, transforming the intra-frame prediction value to obtain the intra-frame transformation coefficient, and obtaining the inter-frame residual coefficient based on the intra-frame transformation coefficient.

In this implementation, inter-frame residual coefficients and intra-frame residual coefficients of attribute information of the node of the target layer are determined; Based on the rate-distortion optimization algorithm, a first generation value corresponding to the inter-frame residual coefficient and a second generation value corresponding to the intra-frame residual coefficient are determined; based on the first generation value and the second generation value, a second prediction residual value is determined. In this way, for the middle-level node, the cost of using intra-frame prediction and inter-frame prediction can be determined through the rate-distortion optimization algorithm, so as to select a suitable prediction method for prediction processing to obtain the second prediction residual value, so that the attribute prediction residual is small and the coding efficiency is high.

Optionally, when the first generation value is less than the second generation value, the second prediction residual value is the inter-frame residual coefficient; or,

When the second generation value is less than the first generation value, the second prediction residual value is the intra-frame residual coefficient; or

When the first generation value is equal to the second generation value, the second prediction residual value is the inter-frame residual coefficient or the intra-frame residual coefficient.

It should be noted that the first generation value being equal to the second generation value indicates that the cost of encoding the intra-frame residual coefficient is the same as the cost of encoding the inter-frame residual coefficient, and the intra-frame prediction mode or the inter-frame prediction mode can be selected.

In this implementation, when the first cost value is less than the second cost value, the second prediction residual value is the inter-frame residual coefficient; when the second cost value is less than the first cost value, the second prediction residual value is the intra-frame residual coefficient; when the first cost value is equal to the second cost value, the second prediction residual value is the inter-frame residual coefficient or the intra-frame residual coefficient. Thus, the prediction method with a smaller cost value can be selected for prediction, so that the attribute prediction residual is smaller and the coding efficiency is higher.

Optionally, the code stream of the point cloud to be encoded also includes a second encoding result, and the second encoding result is used to represent that the second prediction residual value is the intra-frame residual coefficient or the second prediction residual value is the inter-frame residual coefficient.

Among them, a flag can be used to mark the prediction mode used by the node, and the second encoding result can be the encoding result of the flag. For example, a binary flag flag can be used to mark the prediction mode used by the node, when flag is 1, the representation node uses the intra-frame prediction mode, and the second prediction residual value is the intra-frame residual coefficient, when flag is 0, the representation node uses the inter-frame prediction mode, and the second prediction residual value is the inter-frame residual coefficient; or, when flag is 0, the representation node uses the intra-frame prediction mode, and the second prediction residual value is the intra-frame residual coefficient, when flag is 1, the representation node uses the inter-frame prediction mode, and the second prediction residual value is the inter-frame residual coefficient.

In this embodiment, the code stream of the point cloud to be encoded also includes a second encoding result, and the second encoding result is used to characterize that the second prediction residual value is the intra-frame residual coefficient or the second prediction residual value is the inter-frame residual coefficient, so that the decoding end can determine whether the prediction mode is an intra-frame prediction mode or an inter-frame prediction mode through the second encoding result, and thus select a suitable prediction mode for decoding.

Optionally, when the data type of the reference frame data of the to-be-encoded point cloud is the first data type, the determining of the inter-frame residual coefficients of the attribute information of the node of the target layer includes:

Determine an inter-frame prediction value of the attribute information of the node of the target layer, and transform the inter-frame prediction value to obtain an inter-frame transformation coefficient;

An inter-frame residual coefficient is obtained based on the inter-frame transform coefficient.

The reference frame data of the first data type may be encoded and reconstructed point cloud frame data. The reference frame data of the first data type may include attribute information and geometric information.

In one implementation, the transforming the inter-frame prediction value to obtain the inter-frame transform coefficient may include performing a RAHT transform on the inter-frame prediction value to obtain the inter-frame transform coefficient.

In one implementation, the obtaining of the inter-frame residual coefficient based on the inter-frame transform coefficient may include performing a RAHT transform on the attribute information of the node to obtain the transform coefficient of the node; and subtracting the inter-frame transform coefficient from the transform coefficient of the node to obtain the inter-frame residual coefficient. The inter-frame residual coefficient may also be described as an inter-frame prediction residual coefficient.

In this implementation, when the data type of the reference frame data of the point cloud to be encoded is the first data type, the inter-frame prediction value of the attribute information of the node of the target layer is determined, and the inter-frame prediction value is transformed to obtain the inter-frame transformation coefficient; and the inter-frame residual coefficient is obtained based on the inter-frame transformation coefficient. In this way, when the reference frame data is the encoded and reconstructed point cloud frame data, the inter-frame prediction value is transformed to obtain the inter-frame transformation coefficient, and then the inter-frame residual coefficient is obtained.

Optionally, before determining the inter-frame prediction value of the attribute information of the node of the target layer, the method further includes:

Reconstructing points in a reference frame corresponding to the point cloud to be encoded;

Constructing a prediction tree based on the reconstructed reference frame, wherein the tree structure of the prediction tree is the same as the tree structure of the transformation tree;

The determining the inter-frame prediction value of the attribute information of the node of the target layer includes:

An inter-frame prediction value of attribute information of a node of the target layer is determined based on the prediction tree and the transform tree.

Among them, the reconstruction of the reference frame can be realized by reconstructing the points in the reference frame corresponding to the point cloud to be encoded. For each point, there will be a reconstructed reference point with the same position coordinates. The attribute of this reference point is the attribute sum of the points within a certain range in the reference frame obtained according to a certain rule; the rule is: if a point in the reference frame is closer to the current point of the current frame than the Morton code distance to the next point of the current frame, then this point is determined to be used to reconstruct the current point, the attribute value is accumulated, and the weight is increased by one until the condition is not met, and the next reconstruction point is judged. In addition, the starting and ending distances for judgment are specified. If the distance is too far, it will be considered that the attribute value is greatly different and is not suitable for reconstruction. Specifically: only points in the reference frame whose distance to the first point of the current frame is not greater than 64 will be used for reconstruction. Only points in the reference frame whose distance to the last point of the current frame is not greater than 64 will be used for reconstruction. The operation of constructing a prediction tree from bottom to top can be performed on the reconstructed reference frame, and the tree structure of the constructed prediction tree is the same as the tree structure of the transformation tree.

In one embodiment, the inter-frame prediction value of the attribute information of the node of the target layer determined based on the prediction tree and the transform tree may include: obtaining the attribute values of the nodes of each layer and the corresponding attribute values of the nodes of each layer of the reconstructed reference frame through the prediction tree and the transform tree; if the attribute value of the child node at the corresponding position of the reference frame is not 0 and is within 20% to 250% of the attribute value of the parent node of the current frame, then the attribute value of the child node of the reference frame is used as the inter-frame prediction value of the attribute information of the child node at the corresponding position of the current frame; otherwise, the intra-frame prediction value of the current frame is used as the inter-frame prediction value of the attribute information.

In this implementation, the points in the reference frame corresponding to the to-be-encoded point cloud are reconstructed; a prediction tree is constructed based on the reconstructed reference frame, and the tree structure of the prediction tree is the same as the tree structure of the transform tree; and the inter-frame prediction value of the attribute information of the node of the target layer is determined based on the prediction tree and the transform tree. In the case of encoded and reconstructed point cloud frame data, the inter-frame prediction value of the attribute information of the node of the target layer is determined by the prediction tree constructed by the reconstructed reference frame.

Optionally, the reference frame data of the first data type is encoded and reconstructed point cloud frame data.

Optionally, when the data type of the reference frame data of the to-be-encoded point cloud is the second data type, the determining of the inter-frame residual coefficients of the attribute information of the node of the target layer includes:

If there is a transformation coefficient of a reference node having the same position as the node of the target layer in the reference frame of the point cloud to be encoded, the transformation coefficient of the reference node having the same position is determined as the inter-frame transformation coefficient of the attribute information of the node;

An inter-frame residual coefficient of the attribute information of the node is determined based on the transformation coefficient of the node and the inter-frame transformation coefficient of the attribute information of the node.

The determining of the inter-frame residual coefficient of the attribute information of the node based on the transformation coefficient of the node and the inter-frame transformation coefficient of the attribute information of the node may include performing RAHT transformation on the attribute information of the node to obtain the transformation coefficient of the node; and subtracting the inter-frame transformation coefficient of the attribute information of the node from the transformation coefficient of the node to obtain the inter-frame residual coefficient. The inter-frame residual coefficient may also be described as an inter-frame prediction residual coefficient.

In this implementation, if there is a transformation coefficient of a reference node having the same position as the node of the target layer in the reference frame of the point cloud to be encoded, the transformation coefficient of the reference node having the same position is determined as the inter-frame transformation coefficient of the attribute information of the node; based on the transformation coefficient of the node and the inter-frame transformation coefficient of the attribute information of the node, the inter-frame residual coefficient of the attribute information of the node is determined. In this way, when the reference frame data is the reconstructed transformation coefficient of the encoded point cloud frame, the inter-frame transformation coefficient can be directly obtained through the reference frame data, and then the inter-frame residual coefficient can be obtained.

Optionally, the reference frame data of the second data type is reconstructed transformation coefficients of an encoded point cloud frame.

Optionally, the code stream of the point cloud to be encoded further includes a third encoding result, and the third encoding result is an encoding result of at least one of the first preset distance and the second preset distance.

Referring to FIG. 6 , FIG. 6 is a flow chart of a point cloud decoding processing method provided in an embodiment of the present application, which can be applied to a decoding end. As shown in FIG. 6 , the point cloud decoding processing method includes the following steps:

Step 201, determining a target distance between a target layer and a root node of a transform tree corresponding to a point cloud to be decoded;

Step 202: Decode the first encoding result in the code stream of the point cloud to be decoded to obtain a target prediction residual value;

Step 203: acquiring attribute information of the nodes of the target layer based on the target prediction residual value and the target distance;

Wherein, when the target distance is less than or equal to the first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or

In a case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or

When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer.

Optionally, when the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer based on a rate-distortion optimization algorithm.

Optionally, the target prediction residual value includes the second prediction residual value, and the code stream of the point cloud to be decoded also includes a second encoding result, and the second encoding result is used to characterize that the second prediction residual value is an intra-frame residual coefficient or the second prediction residual value is an inter-frame residual coefficient; the acquiring the attribute information of the node of the target layer based on the target prediction residual value and the target distance includes:

When the target distance is greater than the first preset distance and less than or equal to the second preset distance, decoding the second encoding result;

In the case where it is determined that the second encoding result represents that the second prediction residual value is an intra-frame residual coefficient, determining the attribute information of the node of the target layer based on the intra-frame transform coefficient of the node of the target layer and the second prediction residual value; or

When it is determined that the second encoding result represents that the second prediction residual value is an inter-frame residual coefficient, attribute information of the node of the target layer is determined based on the inter-frame transform coefficient of the node of the target layer and the second prediction residual value.

Optionally, when the data type of the reference frame data of the point cloud to be decoded is the first data type, determining the attribute information of the node of the target layer based on the inter-frame transform coefficient of the node of the target layer and the second prediction residual value includes:

Determine an inter-frame prediction value of the attribute information of the node of the target layer, and transform the inter-frame prediction value to obtain an inter-frame transformation coefficient;

The attribute information of the node of the target layer is determined based on the inter-frame transform coefficient and the second prediction residual value.

Optionally, before determining the inter-frame prediction value of the attribute information of the node of the target layer, the method further includes:

Reconstructing points in a reference frame corresponding to the point cloud to be decoded;

Constructing a prediction tree based on the reconstructed reference frame, wherein the tree structure of the prediction tree is the same as the tree structure of the transformation tree;

The determining of the inter-frame prediction value of the attribute information of the node of the target layer includes:

An inter-frame prediction value of attribute information of a node of the target layer is determined based on the prediction tree and the transform tree.

Optionally, the reference frame data of the first data type is decoded and reconstructed point cloud frame data.

Optionally, when the data type of the reference frame data of the point cloud to be decoded is the second data type, determining the attribute information of the node of the target layer based on the inter-frame transform coefficient of the node of the target layer and the second prediction residual value includes:

If it is determined that there is a transformation coefficient of a reference node having the same position as the node of the target layer in the reference frame of the point cloud to be decoded, the transformation coefficient of the reference node having the same position is determined as the inter-frame transformation coefficient of the attribute information of the node;

The attribute information of the node is obtained based on the second prediction residual value and the inter-frame transform coefficient of the attribute information of the node.

Optionally, the reference frame data of the second data type is reconstructed transform coefficients of a decoded point cloud frame.

Optionally, the code stream of the point cloud to be decoded further includes a third encoding result, and the method further includes:

The third encoding result is decoded to obtain at least one of the first preset distance and the second preset distance.

It should be noted that this embodiment is an implementation of the decoding side corresponding to the embodiment shown in Figure 4. Its specific implementation can refer to the relevant description of the embodiment shown in Figure 4. In order to avoid repeated description, this embodiment will not be repeated, and the same beneficial effects can be achieved.

The point cloud encoding processing method and the point cloud decoding processing method of the embodiment of the present application are described below through two specific embodiments:

Embodiment 1:

This embodiment is executed by the encoding end.

This embodiment proposes a RAHT inter-frame prediction scheme based on dual thresholds, by introducing two syntax elements thrh1 (i.e., the first preset distance) and thrh2 (i.e., the second preset distance) in the parameter set (the parameter set can be an attribute parameter set APS, or an attribute data unit ADU, or other parameter sets), and further subdividing the constructed transform tree into three levels: upper layer, middle layer, and lower layer, and using different prediction modes for nodes in different layers. Where thrh1 and thrh2 are the number of layers from the root node, respectively, and generally speaking, thrh1<thrh2. Let the distance from the current node layer to the root node be lvl (i.e., the target distance).

The specific encoding method is described as follows:

Step (11): Reorder the point cloud and construct an N-layer transformation tree structure for the reordered point cloud data based on the geometric distance. A bottom-up construction method is adopted.

In the process of constructing the transformation tree, it is necessary to generate corresponding Morton code information, attribute information and weight information for the merged nodes.

Step (12): Parse the tree from top to bottom. Starting from the root node, upsampling prediction and regional adaptive hierarchical transform (RAHT) are performed layer by layer.

When predicting the nodes of the transform tree, two prediction methods can be selected: intra-frame prediction or inter-frame prediction. As shown in Figure 7, the specific selection method is as follows:

1) For nodes in the upper layer (lvl<=thrh1): Since the nodes in the upper layer of the transform tree are larger and contain more points, even if there is movement between two frames, there will basically not be much change in these upper layer nodes. At this time, if inter-frame prediction is used, the inter-frame attribute prediction residual may be smaller than the intra-frame attribute prediction residual, which means that the use of inter-frame prediction in the upper layer will have a better effect.

Therefore, the inter-frame prediction mode is used by default in the upper layer.

2): For the nodes in the middle layer (thrh1<lvl<=thrh2): some of these nodes in the middle layer may be suitable for inter-frame prediction, and some may be suitable for intra-frame prediction. This embodiment uses the rate-distortion optimization (RDO) algorithm to select the appropriate prediction technology.

The RDO algorithm is used to determine the cost of using intra-frame and inter-frame prediction. When the cost of using inter-frame prediction is less than the cost of using intra-frame prediction, the current node uses the inter-frame prediction mode, otherwise it uses the intra-frame prediction mode. When using a binary flag, it indicates the prediction mode used by the current node. For example, 1 indicates the use of inter-frame prediction mode, and 0 indicates the use of intra-frame prediction mode. And vice versa.

3) For nodes in the lower layer (lvl>thrh2): Since the lower layer blocks are small and contain fewer points, when there is a certain amount of motion between two frames, the number of points contained in each block may change greatly, which is not suitable for inter-frame prediction.

Therefore, the intra prediction mode is used by default in the lower layer.

Among them, for the upper-layer nodes in 1) and the middle-layer nodes in 2), the inter-frame prediction used needs to select different prediction methods according to different reference frame data types.

(a): When the reference frame data is the point cloud frame information that has been encoded and reconstructed, including attribute information and geometric information, the same bottom-up transformation tree construction operation as the current frame can be performed, and the reference frame node attribute value corresponding to the current frame node can be generated, which can be used as the inter-frame prediction value.

The specific process is as follows:

(a1): Reference frame reconstruction: Merge the points in the reference frame within a certain range of the current point into one point, add the attributes, and take the coordinates of the current point. Then, the same prediction tree construction and analysis can be performed. The specific process is as follows:

For each point, there is a reconstructed reference point with the same position coordinates. The attributes of this reference point are the sum of the attributes of the points within a certain range in the reference frame obtained according to certain rules.

The rule is: if the distance of a point in the reference frame to the current point of the current frame is closer than the Morton code distance to the next point of the current frame, then this point is determined to be used to reconstruct the current point, the attribute value is accumulated, and the weight is increased by one. Until the condition is not met, continue to determine the next reconstruction point.

In addition, the start and end distances are specified. If the distance is too far, it will be considered that the attribute values are too different and not suitable for reconstruction.

Only points in the reference frame whose distance to the first point of the current frame is no greater than 64 will be used for reconstruction.

Only points in the reference frame whose distance to the last point of the current frame is no greater than 64 will be used for reconstruction.

(a2): Bottom-up tree construction: The bottom-up prediction tree construction operation is performed on the reconstructed reference frame and the current frame to obtain a prediction tree with the same structure, generating the attribute values and weight values of the nodes at each layer of the current frame and the corresponding predicted attribute values and weight values of the nodes at each layer of the reconstructed reference frame.

(a3): Top-down parse tree:

(a31): For the upper layer nodes, when the number of layers from the current layer to the root node is less than or equal to thrh1, the inter-frame prediction method is used:

First, determine whether the current 2*2*2 node is predicted:

If the current node is a root node, no prediction is performed, and the attribute information of the node is directly subjected to RAHT transformation to obtain the DC coefficient and AC coefficient. The DC coefficient prediction residual is used to replace the DC coefficient obtained by direct transformation.

The DC coefficient prediction residual is: the DC coefficient of the root node of the current frame minus the DC coefficient of the root node of the reference frame.
DC _residual =DC _current -DC _reference

If the current node is not the root node, it is necessary to determine whether to predict the current eight child nodes: For details, please refer to The second step of the intra-frame RAHT technique is described in the content description of the GPCC codec. If no prediction is performed, the original attribute value is directly subjected to RAHT transformation to obtain AC coefficients, which are then quantized and entropy coded.

If it is determined that the current 2*2*2 node is to be predicted, the related method of the third step weighted prediction of the intra-frame RAHT technology in the content description part of the GPCC codec is used to obtain the intra-frame attribute prediction value of each placeholder sub-node of the current 2*2*2 node.

Then, a judgment is made for the placeholder child nodes of each 2*2*2 node: if the attribute value of the child node at the corresponding position in the reference frame is not 0 and is between 20% and 250% of the attribute value of the parent node in the current frame, the attribute value of the child node in the reference frame is used as the attribute prediction value of the child node at the corresponding position in the current frame; otherwise, the intra-frame prediction value of the current frame is used as the attribute prediction value.

Then, the predicted attribute value Attr _predicted obtained by the 2*2*2 node is subjected to RAHT transformation to obtain the AC coefficient of the predicted value. The AC coefficient AC _current obtained by the current frame transformation is subtracted from the predicted AC coefficient AC _predicted to obtain the AC coefficient prediction residual AC _residual , and the residual is quantized and encoded. As shown below:
AC _residual =AC _current -AC _predicted
AC _peedicted =Transform(Attr _peedicted )
Attr _predicted =Attr _{predicted_inter} ? Attr _{predicted_inter} :Attr _{predicted_intra}

(a32): For middle-level nodes, when the number of layers from the current layer to the root node is greater than thrh1 and less than or equal to thrh2, the RDO algorithm is used to decide whether to use inter-frame prediction or intra-frame prediction:

First, determine whether the current 2*2*2 node is predicted: For details, please refer to the description of the second step of the intra-frame RAHT technology in the content description of GPCC encoding and decoding. If no prediction is performed, the original attribute value is directly RAHT transformed to obtain the transformation coefficient, and the transformation coefficient is quantized and entropy encoded.

If it is determined that the current 2*2*2 node is to be predicted, the related method of the third step weighted prediction of the intra-frame RAHT technology in the content description part of the GPCC codec is used to obtain the intra-frame attribute prediction value of each placeholder sub-node of the current 2*2*2 node.

Then, for each 2*2*2 node, the RDO algorithm is used to determine which prediction mode to use. The specific process is as follows:

The inter-frame prediction value and the intra-frame prediction value are respectively subjected to RAHT transformation to obtain corresponding transformation coefficients.

The inter-frame prediction residual coefficient and the intra-frame prediction residual coefficient are respectively subtracted from the true value RAHT transformation coefficient.

According to the RDO algorithm, it is determined whether to use the inter-frame prediction residual coefficient or the intra-frame prediction residual coefficient:

Cost_inter and Cost_intra using intra-frame and inter-frame residual coefficients are calculated respectively, and the cost (Cost) of encoding intra-frame or inter-frame residual coefficients is obtained by the following formula:
Cost＝D+λ*R

Where D (distortion) is the sum of absolute errors (SAD) between the reconstructed attribute and the original attribute obtained after using the prediction residual coefficient (other methods may also be used).

R (rate: bit rate) is the number of bits required to encode the prediction residual coefficients.

The value of λ depends on QP and is calculated by the following formula: (other formulas are also possible)
λ^2＝0.85*2^((QP-4)/3)

If Cost_inter is less than Cost_intra, the inter-frame prediction residual is used, otherwise the intra-frame prediction residual is used. A flag is used to mark this. For example, 1 marks the use of inter-frame prediction residual, and 0 marks the use of intra-frame prediction residual. And vice versa.

The determined AC residual coefficients are quantized and entropy coded.

(a33): For lower layer nodes, when the number of layers from the current layer to the root node is greater than thrh2, the intra-frame prediction method is used:

First, determine whether the current 2*2*2 node is predicted: For details, please refer to the description of the second step of the intra-frame RAHT technology in the content description of GPCC encoding and decoding. If no prediction is performed, the original attribute value is directly RAHT transformed to obtain the transformation coefficient, and the transformation coefficient is quantized and entropy encoded.

If it is determined that the current 2*2*2 node is to be predicted, the related method of the third step weighted prediction of the intra-frame RAHT technology in the content description part of the GPCC codec is used to obtain the intra-frame attribute prediction value of each placeholder sub-node of the current 2*2*2 node.

Then, the intra-frame prediction value is used as the attribute prediction value, and the original attribute value and the intra-frame attribute prediction value are subjected to RAHT transformation and difference, and the residual is quantized and entropy coded.

(b): When the reference frame data is the reconstructed transform coefficient of an already encoded and completed point cloud frame, the transform coefficient can be directly used as the transform coefficient for inter-frame prediction and compared with the transform coefficient of the intra-frame prediction attribute value in the current frame node.

The specific encoding method is as follows:

(b1): For the upper layer nodes, when the number of layers from the current layer to the root node is less than or equal to thrh1, the inter-frame prediction method is used:

Perform RAHT transform on the attribute information of the current node to obtain the transform coefficient, and then determine whether there is a transform coefficient of a reference node with the same position as the current node to be encoded in the reference frame. If so, use the transform coefficient as the inter-frame prediction value of the transform coefficient, and subtract it from the transform coefficient of the current node in the current frame to obtain the transform residual coefficient for subsequent quantization and entropy coding.

(b2): For middle-level nodes, when the number of layers from the current layer to the root node is greater than thrh1 and less than or equal to thrh2, the RDO algorithm is used to decide whether to use inter-frame prediction or intra-frame prediction:

First, determine whether the current 2*2*2 node is predicted: For details, please refer to the description of the second step of the intra-frame RAHT technology in the content description of GPCC codec. If no prediction is performed, the original attribute value is directly RAHT transformed to obtain the transformation coefficient, and the transformation coefficient is quantized and entropy encoded.

If it is determined to make a prediction, the related method of the third step weighted prediction of the intra-frame RAHT technology in the content description part of the GPCC codec is used to obtain the intra-frame attribute prediction value of each placeholder sub-node of the current 2*2*2 node.

Then, for each 2*2*2 node, the RDO algorithm is used to determine which prediction mode to use. The specific process is as follows:

The original attribute value and the intra-frame prediction value of the current node are transformed respectively to obtain the corresponding transformation coefficients, and the intra-frame prediction residual coefficients are obtained by difference.

If there is a transform coefficient at the same position as the current node in the reference frame, it is used as the inter-frame prediction value of the transform coefficient, and the inter-frame prediction residual coefficient is obtained by subtracting it from the actual RAHT transform coefficient of the current node. If it does not exist, the intra-frame prediction residual coefficient is directly used, and subsequent quantization and entropy coding are performed.

According to the RDO algorithm, it is determined whether to use the inter-frame prediction residual coefficient or the intra-frame prediction residual coefficient:

Cost_inter and Cost_intra using intra-frame and inter-frame residual coefficients are calculated respectively, and the cost (Cost) of encoding intra-frame or inter-frame residual coefficients is obtained by the following formula:
Cost＝D+λ*R

Where D (distortion) is the sum of absolute errors (SAD) between the reconstructed attribute and the original attribute obtained after using the prediction residual coefficient (other methods may also be used).

R (rate: bit rate) is the number of bits required to encode the prediction residual coefficients.

The value of λ depends on QP and is calculated by the following formula: (other formulas are also possible)

If Cost_inter is less than Cost_intra, the inter-frame prediction residual is used, otherwise the intra-frame prediction residual is used. A flag is used to mark this. For example, 1 flag uses the inter-frame prediction residual, and 0 flag uses the intra-frame prediction residual. And vice versa.

The determined AC residual coefficients are quantized and entropy encoded.

(b3): For lower-layer nodes, when the number of layers from the current layer to the root node is greater than thrh2, the intra-frame prediction method is used (the same as the encoding method of the lower-layer nodes in (a3)):

First, determine whether the current 2*2*2 node is predicted: For details, please refer to the description of the second step of the intra-frame RAHT technology in the content description of GPCC encoding and decoding. If no prediction is performed, the original attribute value is directly RAHT transformed to obtain the transformation coefficient, and the transformation coefficient is quantized and entropy encoded.

If it is determined that the current 2*2*2 node is to be predicted, the related method of the third step weighted prediction of the intra-frame RAHT technology in the content description part of the GPCC codec is used to obtain the intra-frame attribute prediction value of each placeholder sub-node of the current 2*2*2 node.

Then, the intra-frame prediction value is used as the attribute prediction value, and the original attribute value and the intra-frame attribute prediction value are subjected to RAHT transformation and difference, and the residual is quantized and entropy coded.

Embodiment 2:

This embodiment is executed by a decoding end.

Step (21): First, decode the input code stream to obtain syntax elements thrh1 and thrh2.

The quantized coded transform coefficients or coefficient residuals are entropy decoded and dequantized from the input bitstream to obtain the reconstructed transform coefficients or transform residual coefficients. The flag bit of the prediction mode used by the middle-level node is decoded from the input bitstream.

Since attribute decoding comes after geometric decoding, the reconstructed position coordinates of each point are already obtained when the attribute information is decoded.

Step (22): Reorder the point cloud and construct an N-layer transformation tree structure for the reordered point cloud data based on the geometric distance. A bottom-up construction method is adopted.

Step (23): parse the tree from top to bottom. Starting from the root node, perform upsampling prediction and RAHT inverse transformation layer by layer to obtain the reconstructed attribute value of the node.

When predicting the nodes of the transform tree, two prediction methods can be selected: intra-frame prediction or inter-frame prediction. As shown in Figure 8, the specific selection method is as follows:

1) For nodes in the upper layer (lvl<=thrh1):

By default, inter prediction mode is used.

2): For the nodes in the middle layer (thrh1<lvl<=thrh2): determine whether intra-frame prediction or inter-frame prediction is used according to the prediction mode flag obtained by decoding.

The corresponding prediction value obtained according to the flag is added to the residual transform coefficient obtained by decoding and reconstruction to reconstruct the transform coefficient.

3) For nodes in the lower layer (lvl>thrh2): Since the lower layer blocks are small and contain fewer points, when there is a certain amount of motion between two frames, the number of points contained in each block may change greatly, which is not suitable for inter-frame prediction.

Therefore, the intra prediction mode is used by default in the lower layer.

Among them, for the upper-layer nodes in 1) and the middle-layer nodes in 2), the inter-frame prediction used needs to select different prediction methods according to different reference frame data types.

(a): When the reference frame data is the point cloud frame information that has been encoded and reconstructed, including attribute information and geometric information, the same bottom-up transformation tree construction operation can be performed as the current frame, and the reference frame node attribute value corresponding to the current frame node can be generated, which can be used as the inter-frame prediction value. The specific process has been introduced at the encoding end and will not be repeated here.

When the inter-frame prediction value is obtained, it is added to the decoded transform residual coefficient to obtain the reconstructed transform coefficient, and the RAHT inverse transform is performed to obtain the attribute reconstruction value of each child node.

(a1): For the upper layer nodes, when the number of layers from the current layer to the root node is less than or equal to thrh1, the inter-frame prediction method is used:

First, determine whether the current 2*2*2 node is predicted:

If the current node is a root node, the DC coefficient residual of the root node is added to the DC coefficient residual of the reference frame root node to obtain the DC coefficient of the current frame, and the AC coefficient reconstructed by the root node is used, and then the RAHT inverse transform is performed to obtain the attribute reconstruction value of each child node;

If the current node is not the root node, it is necessary to determine whether to predict the current eight child nodes: for details, please refer to the description of the second step of the intra-frame RAHT technology in the content description section of the GPCC codec.

If no prediction is performed, the DC coefficient directly inherits the attribute reconstruction value of the parent node, and uses the reconstructed AC coefficient of the current node to perform RAHT inverse transformation to obtain the attribute reconstruction value of each child node;

If it is determined that the current 2*2*2 node is to be predicted, the related method of the third step weighted prediction of the intra-frame RAHT technology in the content description part of the GPCC codec is used to obtain the intra-frame attribute prediction value of each placeholder sub-node of the current 2*2*2 node.

Then, a judgment is made for the placeholder child nodes of each 2*2*2 node: if the attribute value of the child node at the corresponding position in the reference frame is not 0 and is between 20% and 250% of the attribute value of the parent node in the current frame, the attribute value of the child node in the reference frame is used as the attribute prediction value of the child node at the corresponding position in the current frame; otherwise, the intra-frame prediction value of the current frame is used as the attribute prediction value.

Then, the attribute prediction value obtained by the 2*2*2 node is transformed by RAHT to obtain the AC coefficient of the attribute prediction value. The DC coefficient inherits the attribute reconstruction value of the parent node, and the AC coefficient is the reconstructed AC coefficient residual plus the AC coefficient of the attribute prediction value. Then, the RAHT inverse transformation is performed to obtain the attribute reconstruction value of each child node.

(a2): For middle-level nodes, when the number of layers from the current layer to the root node is greater than thrh1 and less than or equal to thrh2, the flag is used to determine whether to use inter-frame prediction or intra-frame prediction:

First, determine whether the current 2*2*2 node is predicted: For details, please refer to the GPCC codec content description section The second step of the intra-RAHT technique is described in detail.

If no prediction is performed, the DC coefficient directly inherits the attribute reconstruction value of the parent node, and uses the reconstructed AC coefficient of the current node to perform RAHT inverse transformation to obtain the attribute reconstruction value of each child node;

If it is determined that the current 2*2*2 node is to be predicted, the related method of the third step weighted prediction of the intra-frame RAHT technology in the content description part of the GPCC codec is used to obtain the intra-frame attribute prediction value of each placeholder sub-node of the current 2*2*2 node.

Determine whether intra-frame or inter-frame prediction mode is currently used based on the flag.

RAHT transform is performed on the inter-frame prediction value and the intra-frame prediction value respectively to obtain the corresponding AC coefficients.

The DC coefficient inherits the attribute reconstruction value of the parent node, and the AC coefficient is the reconstructed AC coefficient residual plus the AC coefficient corresponding to the flag. Then, the RAHT inverse transform is performed to obtain the attribute reconstruction value of each child node.

(a3): For lower layer nodes, when the number of layers from the current layer to the root node is greater than thrh2, the intra-frame prediction method is used:

First, determine whether the current 2*2*2 node is predicted: For details, please refer to the description of the second step of the intra-frame RAHT technology in the content description section of GPCC encoding and decoding.

If no prediction is performed, the DC coefficient directly inherits the attribute reconstruction value of the parent node, and uses the reconstructed AC coefficient of the current node to perform RAHT inverse transformation to obtain the attribute reconstruction value of each child node;

If it is determined that the current 2*2*2 node is to be predicted, the related method of the third step weighted prediction of the intra-frame RAHT technology in the content description part of the GPCC codec is used to obtain the intra-frame attribute prediction value of each placeholder sub-node of the current 2*2*2 node.

Then, the intra-frame attribute prediction value obtained by the 2*2*2 node is subjected to RAHT transformation to obtain the AC coefficient of the attribute prediction value.

The DC coefficient inherits the attribute reconstruction value of the parent node, and the AC coefficient is the reconstructed AC coefficient residual plus the AC coefficient of the attribute prediction value. After that, the RAHT inverse transform is performed to obtain the attribute reconstruction value of each child node.

(b): When the reference frame data is the reconstructed transform coefficient of an already encoded and completed point cloud frame, the transform coefficient can be directly used as the transform coefficient for inter-frame prediction and compared with the transform coefficient of the intra-frame prediction attribute value in the current frame node.

(b1): For the upper layer nodes, when the number of layers from the current layer to the root node is less than or equal to thrh1, the inter-frame prediction method is used

Determine whether there is a transform coefficient of a reference node with the same position as the current node to be encoded in the reference frame. If so, use the transform coefficient as the inter-frame prediction value of the transform coefficient, and sum it with the reconstructed transform residual coefficient obtained by decoding to obtain the reconstructed transform coefficient of the current node, perform RAHT inverse transform, and obtain the attribute reconstruction value of each child node.

If it does not exist, the decoded reconstruction transform coefficients are directly subjected to RAHT inverse transformation to obtain the attribute reconstruction value of each child node.

(b2): For middle-level nodes, when the number of layers from the current layer to the root node is greater than thrh1 and less than or equal to thrh2, the flag is used to determine whether to use inter-frame prediction or intra-frame prediction:

First, determine whether the current 2*2*2 node is predicted: For details, please refer to the description of the second step of the intra-frame RAHT technology in the content description section of GPCC encoding and decoding.

If no prediction is performed, the DC coefficient directly inherits the attribute reconstruction value of the parent node, and uses the reconstructed AC coefficient of the current node to perform RAHT inverse transformation to obtain the attribute reconstruction value of each child node;

If it is determined that the current 2*2*2 node is to be predicted, the related method of the third step weighted prediction of the intra-frame RAHT technology in the content description part of the GPCC codec is used to obtain the intra-frame attribute prediction value of each placeholder sub-node of the current 2*2*2 node.

1. Transform the intra-frame prediction value to obtain the transform coefficient of the intra-frame prediction.

2. Determine whether there is a transform coefficient at the same position as the current node in the reference frame. If not, directly use the intra-frame prediction coefficient and sum it with the decoded reconstructed transform residual coefficient to obtain the reconstructed transform coefficient of the current node, perform RAHT inverse transform, and obtain the attribute reconstruction value of each child node.

3. If it exists, determine whether the current prediction mode is intra-frame or inter-frame according to the flag. Get the predicted transform coefficient according to the flag, sum it with the decoded reconstruction transform residual coefficient to get the reconstruction transform coefficient of the current node, perform RAHT inverse transform, and get the attribute reconstruction value of each child node.

(b3): For lower layer nodes, when the number of layers from the current layer to the root node is greater than thrh2, the intra-frame prediction method is used, which is the same as the lower layer decoding method in (a3) and will not be repeated here.

At this point, attribute decoding is complete.

This embodiment proposes a dual-threshold inter-frame prediction method, which divides the octree into three levels: upper level, middle level, and lower level. Different prediction methods are used for different levels according to the node characteristics of different levels. In particular, for the nodes in the middle level, the RDO algorithm is used to determine whether the current node is an intra-frame or inter-frame prediction method. At the same time, the two set thresholds and flags need to be transmitted in the bitstream. The three-layer structure of this embodiment is more flexible and has higher coding efficiency.

Compared with the prior art, this embodiment is more flexible. Considering that some nodes in the middle layer are more suitable for the inter-frame prediction method, a double threshold is used to divide the octree into three levels. For the nodes in the middle layer, the RDO algorithm is used to determine whether to adopt the inter-frame prediction mode, and a flag is used to mark it. This is more flexible and can reduce the bit rate and improve the coding efficiency.

It should be noted that the point cloud coding processing method provided in the embodiment of the present application can be executed by a point cloud coding processing device, or a control module in the point cloud coding processing device for executing the point cloud coding processing method. In the embodiment of the present application, the point cloud coding processing device provided in the embodiment of the present application is described by taking the method for executing the point cloud coding processing by the point cloud coding processing device as an example.

Please refer to FIG. 9 , which is a structural diagram of a point cloud coding processing device provided in an embodiment of the present application. As shown in FIG. 9 , the point cloud coding processing device 300 includes:

The first determination module 301 is used to determine the target distance between the target layer and the root node of the transform tree corresponding to the point cloud to be encoded;

A second determination module 302 is used to determine a target prediction residual value based on the target distance;

The encoding module 303 is used to encode the target prediction residual value to obtain a first encoding result;

The code stream of the point cloud to be encoded includes the first encoding result;

In a case where the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or

When the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or

When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer.

Optionally, when the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer based on a rate-distortion optimization algorithm.

Optionally, the second determining module includes:

A first determining unit, configured to determine an inter-frame residual coefficient and an intra-frame residual coefficient of the attribute information of the node of the target layer when the target distance is greater than the first preset distance and less than or equal to the second preset distance;

A second determining unit, configured to determine a first generation value corresponding to the inter-frame residual coefficient and a second generation value corresponding to the intra-frame residual coefficient based on a rate-distortion optimization algorithm;

a third determining unit, configured to determine a second prediction residual value based on the first generation value and the second generation value;

Among them, the target prediction residual value includes the second prediction residual value.

Optionally, when the first generation value is less than the second generation value, the second prediction residual value is the inter-frame residual coefficient; or,

When the second generation value is less than the first generation value, the second prediction residual value is the intra-frame residual coefficient; or

When the first generation value is equal to the second generation value, the second prediction residual value is the inter-frame residual coefficient or the intra-frame residual coefficient.

Optionally, the code stream of the point cloud to be encoded also includes a second encoding result, and the second encoding result is used to represent that the second prediction residual value is the intra-frame residual coefficient or the second prediction residual value is the inter-frame residual coefficient.

Optionally, the first determining unit includes:

A determination subunit, configured to determine, when the data type of the reference frame data of the to-be-encoded point cloud is the first data type, an inter-frame prediction value of the attribute information of the node of the target layer, and transform the inter-frame prediction value to obtain an inter-frame transformation coefficient;

The acquisition subunit is used to acquire the inter-frame residual coefficient based on the inter-frame transform coefficient.

Optionally, the first determining unit is further configured to:

Reconstructing points in a reference frame corresponding to the point cloud to be encoded;

Constructing a prediction tree based on the reconstructed reference frame, wherein the tree structure of the prediction tree is the same as the tree structure of the transformation tree;

The determining subunit is specifically used for:

An inter-frame prediction value of attribute information of a node of the target layer is determined based on the prediction tree and the transform tree.

Optionally, the reference frame data of the first data type is encoded and reconstructed point cloud frame data.

Optionally, when the data type of the reference frame data of the to-be-encoded point cloud is the second data type, the first determining unit is specifically configured to:

If there is a transformation coefficient of a reference node having the same position as the node of the target layer in the reference frame of the point cloud to be encoded, the transformation coefficient of the reference node having the same position is determined as the inter-frame transformation coefficient of the attribute information of the node;

An inter-frame residual coefficient of the attribute information of the node is determined based on the transformation coefficient of the node and the inter-frame transformation coefficient of the attribute information of the node.

Optionally, the reference frame data of the second data type is reconstructed transformation coefficients of an encoded point cloud frame.

Optionally, the code stream of the point cloud to be encoded further includes a third encoding result, and the third encoding result is an encoding result of at least one of the first preset distance and the second preset distance.

The point cloud coding processing device in the embodiment of the present application can be a device, a device or electronic device with an operating system, or a component, integrated circuit, or chip in a terminal. The device or electronic device can be a mobile terminal or a non-mobile terminal. Exemplarily, the mobile terminal can include but is not limited to the types of terminals listed above, and the non-mobile terminal can be a server, a network attached storage (Network Attached Storage, NAS), a personal computer (personal computer, PC), a television (television, TV), a teller machine or a self-service machine, etc., which is not specifically limited in the embodiment of the present application.

The point cloud coding processing device provided in the embodiment of the present application can implement each process implemented by the method embodiment of Figure 5 and achieve the same technical effect. To avoid repetition, it will not be repeated here.

It should be noted that the point cloud decoding processing method provided in the embodiment of the present application can be executed by a point cloud decoding processing device, or a control module in the point cloud decoding processing device for executing the point cloud decoding processing method. In the embodiment of the present application, the point cloud decoding processing device provided in the embodiment of the present application is described by taking the method for executing the point cloud decoding processing by the point cloud decoding processing device as an example.

Please refer to FIG. 10 , which is a structural diagram of a point cloud decoding processing device provided in an embodiment of the present application. As shown in FIG. 10 , the point cloud decoding processing device 400 includes:

A determination module 401 is used to determine a target distance between a target layer and a root node of a transform tree corresponding to a point cloud to be decoded;

A decoding module 402, configured to decode a first encoding result in a bitstream of a point cloud to be decoded, and obtain a target prediction residual value;

An acquisition module 403 is used to acquire attribute information of the node of the target layer based on the target prediction residual value and the target distance;

Wherein, when the target distance is less than or equal to the first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or

In a case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or

When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer.

Optionally, when the target distance is greater than the first preset distance and less than or equal to the second preset distance, The target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer based on a rate-distortion optimization algorithm.

Optionally, the target prediction residual value includes the second prediction residual value, the code stream of the point cloud to be decoded also includes a second encoding result, and the second encoding result is used to represent that the second prediction residual value is an intra-frame residual coefficient or the second prediction residual value is an inter-frame residual coefficient; the acquisition module includes:

a decoding unit, configured to decode the second encoding result when the target distance is greater than the first preset distance and less than or equal to a second preset distance;

A determination unit, configured to determine attribute information of a node of the target layer based on an intra-frame transform coefficient of the node of the target layer and the second prediction residual value when it is determined that the second encoding result represents that the second prediction residual value is an intra-frame residual coefficient; or

When it is determined that the second encoding result represents that the second prediction residual value is an inter-frame residual coefficient, attribute information of the node of the target layer is determined based on the inter-frame transform coefficient of the node of the target layer and the second prediction residual value.

Optionally, the determining unit includes:

A first determination subunit is used to determine the inter-frame prediction value of the attribute information of the node of the target layer when the data type of the reference frame data of the to-be-decoded point cloud is the first data type, and to transform the inter-frame prediction value to obtain an inter-frame transformation coefficient;

The second determination subunit is used to determine the attribute information of the node of the target layer based on the inter-frame transformation coefficient and the second prediction residual value.

Optionally, the determining unit is further configured to:

Reconstructing points in a reference frame corresponding to the point cloud to be decoded;

Constructing a prediction tree based on the reconstructed reference frame, wherein the tree structure of the prediction tree is the same as the tree structure of the transformation tree;

The first determining subunit is specifically used for:

An inter-frame prediction value of attribute information of a node of the target layer is determined based on the prediction tree and the transform tree.

Optionally, the reference frame data of the first data type is decoded and reconstructed point cloud frame data.

Optionally, when the data type of the reference frame data of the to-be-decoded point cloud is the second data type, when determining that the second encoding result represents that the second prediction residual value is an inter-frame residual coefficient, the determining unit is specifically configured to:

If it is determined that there is a transformation coefficient of a reference node having the same position as the node of the target layer in the reference frame of the point cloud to be decoded, the transformation coefficient of the reference node having the same position is determined as the inter-frame transformation coefficient of the attribute information of the node;

The attribute information of the node is obtained based on the second prediction residual value and the inter-frame transform coefficient of the attribute information of the node.

Optionally, the reference frame data of the second data type is reconstructed transform coefficients of a decoded point cloud frame.

Optionally, the code stream of the point cloud to be decoded further includes a third encoding result, and the decoding module is further used to:

The third encoding result is decoded to obtain at least one of the first preset distance and the second preset distance.

The point cloud decoding processing device in the embodiment of the present application can be a device, a device or electronic device with an operating system, or a component, integrated circuit, or chip in a terminal. The device or electronic device can be a mobile terminal or a non-mobile terminal. Exemplarily, the mobile terminal can include but is not limited to the types of terminals listed above, and the non-mobile terminal can be a server, a network attached storage (Network Attached Storage, NAS), a personal computer (personal computer, PC), a television (television, TV), a teller machine or a self-service machine, etc., which is not specifically limited in the embodiment of the present application.

The point cloud decoding processing device provided in the embodiment of the present application can implement the various processes implemented by the method embodiment of Figure 6 and achieve the same technical effect. To avoid repetition, it will not be repeated here.

Optionally, as shown in FIG11 , the embodiment of the present application further provides a communication device 500, including a processor 501 and a memory 502, wherein the memory 502 stores a program or instruction that can be run on the processor 501. For example, when the communication device 500 is an encoding end device, the program or instruction is executed by the processor 501 to implement the various steps of the above-mentioned point cloud encoding processing method embodiment, and can achieve the same technical effect. When the communication device 500 is a decoding end device, the program or instruction is executed by the processor 501 to implement the various steps of the above-mentioned point cloud decoding processing method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

The embodiment of the present application also provides a terminal, including a processor and a communication interface, wherein the processor is used to: determine the target distance between the target layer and the root node of the transform tree corresponding to the point cloud to be encoded; determine the target prediction residual value based on the target distance; encode the target prediction residual value to obtain a first encoding result; wherein the code stream of the point cloud to be encoded includes the first encoding result; when the target distance is less than or equal to the first preset distance, the target prediction residual value includes a first prediction residual value obtained by inter-frame prediction of the attribute information of the node of the target layer; or when the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer; or when the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by intra-frame prediction of the attribute information of the node of the target layer. This terminal embodiment corresponds to the above-mentioned point cloud encoding processing method embodiment, and each implementation process and implementation method of the above-mentioned point cloud encoding processing method embodiment can be applied to this terminal embodiment and can achieve the same technical effect.

An embodiment of the present application also provides a terminal, including a processor and a communication interface, wherein the processor is used to: determine a target distance between a target layer and a root node of a transform tree corresponding to a point cloud to be decoded; decode a first encoding result in a code stream of a point cloud to be decoded to obtain a target prediction residual value; obtain attribute information of the nodes of the target layer based on the target prediction residual value and the target distance; wherein, when the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the nodes of the target layer; or when the target distance is greater than the first preset distance and less than or equal to a second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the nodes of the target layer; or when the target distance is greater than the second preset distance, the target prediction residual value includes The third prediction residual value obtained by intra-frame prediction of the attribute information of the node of the target layer. This terminal embodiment corresponds to the above-mentioned point cloud decoding processing method embodiment, and each implementation process and implementation method of the above-mentioned point cloud decoding processing method embodiment can be applied to this terminal embodiment and can achieve the same technical effect.

Specifically, FIG12 is a schematic diagram of the hardware structure of a terminal implementing an embodiment of the present application.

The terminal 600 includes but is not limited to: a radio frequency unit 601, a network module 602, an audio output unit 603, an input unit 604, a sensor 605, a display unit 606, a user input unit 607, an interface unit 608, a memory 609 and at least some of the components of a processor 610.

Those skilled in the art will appreciate that the terminal 600 may also include a power source (such as a battery) for supplying power to each component, and the power source may be logically connected to the processor 610 through a power management system, so as to implement functions such as managing charging, discharging, and power consumption management through the power management system. The terminal structure shown in FIG12 does not constitute a limitation on the terminal, and the terminal may include more or fewer components than shown in the figure, or combine certain components, or arrange components differently, which will not be described in detail here.

It should be understood that in the embodiment of the present application, the input unit 604 may include a graphics processing unit (GPU) 6041 and a microphone 6042, and the graphics processor 6041 processes the image data of the static picture or video obtained by the image capture device (such as a camera) in the video capture mode or the image capture mode. The display unit 606 may include a display panel 6061, and the display panel 6061 may be configured in the form of a liquid crystal display, an organic light emitting diode, etc. The user input unit 607 includes a touch panel 6071 and at least one of other input devices 6072. The touch panel 6071 is also called a touch screen. The touch panel 6071 may include two parts: a touch detection device and a touch controller. Other input devices 6072 may include, but are not limited to, a physical keyboard, function keys (such as a volume control key, a switch key, etc.), a trackball, a mouse, and a joystick, which will not be repeated here.

In the embodiment of the present application, after receiving downlink data from the network side device, the RF unit 601 can transmit the data to the processor 610 for processing; in addition, the RF unit 601 can send uplink data to the network side device. Generally, the RF unit 601 includes but is not limited to an antenna, an amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, etc.

The memory 609 can be used to store software programs or instructions and various data. The memory 609 may mainly include a first storage area for storing programs or instructions and a second storage area for storing data, wherein the first storage area may store an operating system, an application program or instruction required for at least one function (such as a sound playback function, an image playback function, etc.), etc. In addition, the memory 609 may include a volatile memory or a non-volatile memory, or the memory 609 may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), a static random access memory (SRAM), a dynamic random access memory (DRAM), a synchronous dynamic random access memory (SDRAM), a double data rate synchronous dynamic random access memory (DDRSDRAM), an enhanced synchronous dynamic random access memory (ESDRAM), a synchronous link dynamic random access memory (SLDRAM) and a direct RAM bus random access memory (DRRAM). The memory x09 in the embodiment of the present application includes but Without limitation, these and any other suitable types of memory.

The processor 610 may include one or more processing units; optionally, the processor 610 integrates an application processor and a modem processor, wherein the application processor mainly processes operations related to an operating system, a user interface, and application programs, and the modem processor mainly processes wireless communication signals, such as a baseband processor. It is understandable that the modem processor may not be integrated into the processor 610.

Wherein, when the terminal is a coding terminal:

The processor 610 is configured to:

Determine the target distance between the target layer and the root node of the transform tree corresponding to the point cloud to be encoded;

Determining a target prediction residual value based on the target distance;

Encoding the target prediction residual value to obtain a first encoding result;

The code stream of the point cloud to be encoded includes the first encoding result;

In a case where the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or

In a case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or

When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer.

Optionally, when the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer based on a rate-distortion optimization algorithm.

Optionally, the processor 610 is specifically configured to:

When the target distance is greater than the first preset distance and less than or equal to the second preset distance, determine the inter-frame residual coefficient and the intra-frame residual coefficient of the attribute information of the node of the target layer;

Determine a first generation value corresponding to the inter-frame residual coefficient and a second generation value corresponding to the intra-frame residual coefficient based on a rate-distortion optimization algorithm;

determining a second prediction residual value based on the first generation value and the second generation value;

Among them, the target prediction residual value includes the second prediction residual value.

Optionally, when the first generation value is less than the second generation value, the second prediction residual value is the inter-frame residual coefficient; or,

When the second generation value is less than the first generation value, the second prediction residual value is the intra-frame residual coefficient; or

When the first generation value is equal to the second generation value, the second prediction residual value is the inter-frame residual coefficient or the intra-frame residual coefficient.

Optionally, the code stream of the point cloud to be encoded further includes a second encoding result, and the second encoding result is used to represent the The second prediction residual value is the intra-frame residual coefficient or the second prediction residual value is the inter-frame residual coefficient.

Optionally, when the data type of the reference frame data of the to-be-encoded point cloud is the first data type, the processor 610 is specifically configured to:

Determine an inter-frame prediction value of the attribute information of the node of the target layer, and transform the inter-frame prediction value to obtain an inter-frame transformation coefficient;

An inter-frame residual coefficient is obtained based on the inter-frame transform coefficient.

Optionally, the processor 610 is further configured to:

Reconstructing points in a reference frame corresponding to the point cloud to be encoded;

Constructing a prediction tree based on the reconstructed reference frame, wherein the tree structure of the prediction tree is the same as the tree structure of the transformation tree;

An inter-frame prediction value of attribute information of a node of the target layer is determined based on the prediction tree and the transform tree.

Optionally, the reference frame data of the first data type is encoded and reconstructed point cloud frame data.

Optionally, when the data type of the reference frame data of the to-be-encoded point cloud is the second data type, the processor 610 is specifically configured to:

If there is a transformation coefficient of a reference node having the same position as the node of the target layer in the reference frame of the point cloud to be encoded, the transformation coefficient of the reference node having the same position is determined as the inter-frame transformation coefficient of the attribute information of the node;

An inter-frame residual coefficient of the attribute information of the node is determined based on the transformation coefficient of the node and the inter-frame transformation coefficient of the attribute information of the node.

Optionally, the reference frame data of the second data type is reconstructed transformation coefficients of an encoded point cloud frame.

Optionally, the code stream of the point cloud to be encoded further includes a third encoding result, and the third encoding result is an encoding result of at least one of the first preset distance and the second preset distance.

Wherein, when the terminal is a decoding terminal:

The processor 610 is configured to:

Determine the target distance between the target layer and the root node of the transform tree corresponding to the point cloud to be decoded;

Decode the first encoding result in the bitstream of the point cloud to be decoded to obtain a target prediction residual value;

Acquire attribute information of nodes in the target layer based on the target prediction residual value and the target distance;

Wherein, when the target distance is less than or equal to the first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or

In a case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or

When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer.

Optionally, when the target distance is greater than the first preset distance and less than or equal to the second preset distance, The target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer based on a rate-distortion optimization algorithm.

Optionally, the target prediction residual value includes the second prediction residual value, the code stream of the point cloud to be decoded also includes a second encoding result, and the second encoding result is used to characterize that the second prediction residual value is an intra-frame residual coefficient or the second prediction residual value is an inter-frame residual coefficient; the processor 610 is specifically used to:

When the target distance is greater than the first preset distance and less than or equal to the second preset distance, decoding the second encoding result;

In the case where it is determined that the second encoding result represents that the second prediction residual value is an intra-frame residual coefficient, determining the attribute information of the node of the target layer based on the intra-frame transform coefficient of the node of the target layer and the second prediction residual value; or

When it is determined that the second encoding result represents that the second prediction residual value is an inter-frame residual coefficient, attribute information of the node of the target layer is determined based on the inter-frame transform coefficient of the node of the target layer and the second prediction residual value.

Optionally, when the data type of the reference frame data of the to-be-decoded point cloud is the first data type, the processor 610 is specifically configured to:

Determine an inter-frame prediction value of the attribute information of the node of the target layer, and transform the inter-frame prediction value to obtain an inter-frame transformation coefficient;

The attribute information of the node of the target layer is determined based on the inter-frame transform coefficient and the second prediction residual value.

Optionally, the processor 610 is further configured to:

Reconstructing points in a reference frame corresponding to the point cloud to be decoded;

Constructing a prediction tree based on the reconstructed reference frame, wherein the tree structure of the prediction tree is the same as the tree structure of the transformation tree;

An inter-frame prediction value of attribute information of a node of the target layer is determined based on the prediction tree and the transform tree.

Optionally, the reference frame data of the first data type is decoded and reconstructed point cloud frame data.

Optionally, when the data type of the reference frame data of the to-be-decoded point cloud is the second data type, the processor 610 is specifically configured to:

If it is determined that there is a transformation coefficient of a reference node having the same position as the node of the target layer in the reference frame of the point cloud to be decoded, the transformation coefficient of the reference node having the same position is determined as the inter-frame transformation coefficient of the attribute information of the node;

The attribute information of the node is obtained based on the second prediction residual value and the inter-frame transform coefficient of the attribute information of the node.

Optionally, the reference frame data of the second data type is reconstructed transform coefficients of a decoded point cloud frame.

Optionally, the code stream of the point cloud to be decoded further includes a third encoding result, and the processor 610 is further configured to:

The third encoding result is decoded to obtain at least one of the first preset distance and the second preset distance.

The embodiments of the present application can improve coding efficiency.

Specifically, the terminal of the embodiment of the present application also includes: instructions or programs stored in the memory 609 and executable on the processor 610. The processor 610 calls the instructions or programs in the memory 609 to execute the methods executed by the modules shown in Figure 9 or Figure 10, and achieves the same technical effect. To avoid repetition, it will not be repeated here.

An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the various processes of the above-mentioned point cloud encoding processing method embodiment are implemented, or when the program or instruction is executed by a processor, the various processes of the above-mentioned point cloud decoding processing method embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

The processor is the processor in the terminal described in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk. In some examples, the readable storage medium may be a non-transient readable storage medium.

An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned point cloud encoding processing method embodiment, or to implement the various processes of the above-mentioned point cloud decoding processing method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

The embodiments of the present application further provide a computer program/program product, which is stored in a storage medium. The computer program/program product is executed by at least one processor to implement the various processes of the above-mentioned point cloud encoding processing method or point cloud decoding processing method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

An embodiment of the present application also provides a coding and decoding system, including: an encoding end device and a decoding end device, wherein the encoding end device can be used to execute the steps of the point cloud encoding processing method as described above, and the decoding end device can be used to execute the steps of the point cloud decoding processing method as described above.

It should be noted that, in this article, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises one..." does not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the method and device in the embodiment of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted or combined. In addition, the features described with reference to certain examples may be combined in other examples.

Through the above description of the implementation mode, those skilled in the art can clearly understand that the above embodiment method can be implemented by means of a computer software product plus a necessary general hardware platform, or of course by hardware. The computer software product is stored in a storage medium (such as ROM, RAM, magnetic disk, optical disk, etc.), and includes several instructions for enabling The terminal or network side device executes the methods described in each embodiment of the present application.

The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms of implementation methods without departing from the purpose of the present application and the scope of protection of the claims, and these implementation methods are all within the protection of the present application.

Claims (24)

A point cloud coding processing method, executed by a coding end, includes: Determine the target distance between the target layer and the root node of the transform tree corresponding to the point cloud to be encoded; Determining a target prediction residual value based on the target distance; Encoding the target prediction residual value to obtain a first encoding result; The code stream of the point cloud to be encoded includes the first encoding result; In a case where the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or In a case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer. According to the method according to claim 1, wherein, when the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer based on a rate-distortion optimization algorithm. The method according to claim 1 or 2, wherein determining the target prediction residual value based on the target distance comprises: When the target distance is greater than the first preset distance and less than or equal to the second preset distance, determine the inter-frame residual coefficient and the intra-frame residual coefficient of the attribute information of the node of the target layer; Determine a first generation value corresponding to the inter-frame residual coefficient and a second generation value corresponding to the intra-frame residual coefficient based on a rate-distortion optimization algorithm; determining a second prediction residual value based on the first generation value and the second generation value; Among them, the target prediction residual value includes the second prediction residual value. The method according to claim 3, wherein, when the first generation value is less than the second generation value, the second prediction residual value is the inter-frame residual coefficient; or When the second generation value is less than the first generation value, the second prediction residual value is the intra-frame residual coefficient; or When the first generation value is equal to the second generation value, the second prediction residual value is the inter-frame residual coefficient or the intra-frame residual coefficient. The method according to claim 3 or 4, wherein the code stream of the point cloud to be encoded also includes a second encoding result, and the second encoding result is used to characterize that the second prediction residual value is the intra-frame residual coefficient or the second prediction residual value is the inter-frame residual coefficient. The method according to any one of claims 3 to 5, wherein the reference frame data of the point cloud to be encoded When the data type is the first data type, the inter-frame residual coefficients for determining the attribute information of the node of the target layer include: Determine an inter-frame prediction value of the attribute information of the node of the target layer, and transform the inter-frame prediction value to obtain an inter-frame transformation coefficient; An inter-frame residual coefficient is obtained based on the inter-frame transform coefficient. The method according to claim 6, wherein, before determining the inter-frame prediction value of the attribute information of the node of the target layer, the method further comprises: Reconstructing points in a reference frame corresponding to the point cloud to be encoded; Constructing a prediction tree based on the reconstructed reference frame, wherein the tree structure of the prediction tree is the same as the tree structure of the transformation tree; The determining the inter-frame prediction value of the attribute information of the node of the target layer includes: An inter-frame prediction value of attribute information of a node of the target layer is determined based on the prediction tree and the transform tree. The method according to claim 6 or 7, wherein the reference frame data of the first data type is encoded and reconstructed point cloud frame data. The method according to any one of claims 3 to 5, wherein, when the data type of the reference frame data of the to-be-encoded point cloud is the second data type, the inter-frame residual coefficients of the attribute information of the node of the target layer are determined, comprising: If there is a transformation coefficient of a reference node having the same position as the node of the target layer in the reference frame of the point cloud to be encoded, the transformation coefficient of the reference node having the same position is determined as the inter-frame transformation coefficient of the attribute information of the node; An inter-frame residual coefficient of the attribute information of the node is determined based on the transformation coefficient of the node and the inter-frame transformation coefficient of the attribute information of the node. The method according to claim 9, wherein the reference frame data of the second data type is the reconstructed transform coefficients of the encoded point cloud frame. According to the method according to any one of claims 1-10, wherein the code stream of the point cloud to be encoded also includes a third encoding result, and the third encoding result is an encoding result of at least one of the first preset distance and the second preset distance. A point cloud decoding processing method, executed by a decoding end, includes: Determine the target distance between the target layer and the root node of the transform tree corresponding to the point cloud to be decoded; Decode the first encoding result in the bitstream of the point cloud to be decoded to obtain a target prediction residual value; Acquire attribute information of nodes in the target layer based on the target prediction residual value and the target distance; Wherein, when the target distance is less than or equal to the first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or In the case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer. According to the method according to claim 12, wherein, when the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by predicting the attribute information of the node of the target layer based on a rate-distortion optimization algorithm. The method according to claim 12 or 13, wherein the target prediction residual value includes the second prediction residual value, the code stream of the point cloud to be decoded also includes a second encoding result, and the second encoding result is used to characterize that the second prediction residual value is an intra-frame residual coefficient or the second prediction residual value is an inter-frame residual coefficient; the acquiring the attribute information of the node of the target layer based on the target prediction residual value and the target distance comprises: When the target distance is greater than the first preset distance and less than or equal to the second preset distance, decoding the second encoding result; In a case where it is determined that the second encoding result represents that the second prediction residual value is an intra-frame residual coefficient, determining the attribute information of the node of the target layer based on the intra-frame transform coefficient of the node of the target layer and the second prediction residual value; or When it is determined that the second encoding result represents that the second prediction residual value is an inter-frame residual coefficient, attribute information of the node of the target layer is determined based on the inter-frame transform coefficient of the node of the target layer and the second prediction residual value. The method according to claim 14, wherein, when the data type of the reference frame data of the point cloud to be decoded is the first data type, determining the attribute information of the node of the target layer based on the inter-frame transform coefficient of the node of the target layer and the second prediction residual value comprises: Determine an inter-frame prediction value of the attribute information of the node of the target layer, and transform the inter-frame prediction value to obtain an inter-frame transformation coefficient; The attribute information of the node of the target layer is determined based on the inter-frame transform coefficient and the second prediction residual value. The method according to claim 15, wherein, before determining the inter-frame prediction value of the attribute information of the node of the target layer, the method further comprises: Reconstructing points in a reference frame corresponding to the point cloud to be decoded; Constructing a prediction tree based on the reconstructed reference frame, wherein the tree structure of the prediction tree is the same as the tree structure of the transformation tree; The determining the inter-frame prediction value of the attribute information of the node of the target layer includes: An inter-frame prediction value of attribute information of a node of the target layer is determined based on the prediction tree and the transform tree. The method according to claim 15 or 16, wherein the reference frame data of the first data type is decoded and reconstructed point cloud frame data. The method according to claim 14, wherein, when the data type of the reference frame data of the point cloud to be decoded is the second data type, the inter-frame transform coefficients of the nodes based on the target layer and the second prediction The residual value determines the attribute information of the node of the target layer, including: If it is determined that there is a transformation coefficient of a reference node having the same position as the node of the target layer in the reference frame of the point cloud to be decoded, the transformation coefficient of the reference node having the same position is determined as the inter-frame transformation coefficient of the attribute information of the node; The attribute information of the node is obtained based on the second prediction residual value and the inter-frame transform coefficient of the attribute information of the node. The method according to claim 18, wherein the reference frame data of the second data type is the reconstructed transform coefficients of the decoded point cloud frame. The method according to any one of claims 12 to 19, wherein the code stream of the point cloud to be decoded further includes a third encoding result, and the method further includes: The third encoding result is decoded to obtain at least one of the first preset distance and the second preset distance. A point cloud coding processing device, comprising: A first determination module is used to determine a target distance between a target layer and a root node of a transform tree corresponding to a point cloud to be encoded; A second determination module, used to determine a target prediction residual value based on the target distance; An encoding module, used for encoding the target prediction residual value to obtain a first encoding result; The code stream of the point cloud to be encoded includes the first encoding result; In a case where the target distance is less than or equal to a first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or In a case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer. A point cloud decoding processing device, comprising: A determination module, used to determine a target distance between a target layer and a root node of a transform tree corresponding to a point cloud to be decoded; A decoding module, used for decoding a first encoding result in a bit stream of a point cloud to be decoded to obtain a target prediction residual value; An acquisition module, used for acquiring attribute information of the nodes of the target layer based on the target prediction residual value and the target distance; Wherein, when the target distance is less than or equal to the first preset distance, the target prediction residual value includes a first prediction residual value obtained by performing inter-frame prediction on the attribute information of the node of the target layer; or In a case where the target distance is greater than the first preset distance and less than or equal to the second preset distance, the target prediction residual value includes a second prediction residual value obtained by performing prediction processing on the attribute information of the node of the target layer; or When the target distance is greater than the second preset distance, the target prediction residual value includes a third prediction residual value obtained by performing intra-frame prediction on the attribute information of the node of the target layer. A terminal comprises a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the point cloud encoding processing method as described in any one of claims 1 to 11; or, the program or instruction, when executed by the processor, implements the steps of the point cloud decoding processing method as described in any one of claims 12 to 20. A readable storage medium storing a program or instruction, wherein the program or instruction, when executed by a processor, implements the steps of the point cloud encoding processing method as described in any one of claims 1 to 11, or, when executed by a processor, implements the steps of the point cloud decoding processing method as described in any one of claims 12 to 20.