CN117689557A - Method, system and storage medium for converting histogram panorama to hexagram based on OpenCV - Google Patents

Abstract

The invention discloses a method, system and storage medium for converting a histogram panorama to a hexagonal image based on OpenCV, which belongs to the field of image conversion technology. The method includes: setting the number of converted images and image resolution, and creating a corresponding number according to the number of images. The conversion task is to realize the conversion of the panoramic map to the six-sided image by calling the remap function in OpenCV; define the single-sided image in the six-sided image as the target image, obtain the resolution information of the target image, and store it in the local database Query whether the corresponding mapping data exists for the resolution information. If it exists, the mapping data is directly called and sent to the conversion task; if it does not exist, the new mapping data corresponding to the resolution information is calculated; the new mapping data is sent to the conversion task and Save to local database. The present invention realizes the rapid conversion of panoramic images by using the mapping data calculated by the resolution of the single-sided image as the input of the conversion task.

Description

Method, system and storage medium for converting histogram panorama to hexagram based on OpenCV

Technical field

The present invention relates to the technical field of image conversion, and in particular to a method, system and storage medium for converting a histogram panorama into a hexagonal image based on OpenCV.

Background technique

A panorama is a picture obtained by capturing the image information of the entire scene with a professional camera or rendering it using modeling software. The software is used to stitch the pictures together and played with a special player. It turns a flat photo or a computer modeling picture into a 360-degree panorama. View, used for virtual reality browsing, simulates a two-dimensional plan into a real three-dimensional space and presents it to the viewer.

Panoramas are widely used in virtual scenes. With the development of technology, users are increasingly pursuing the real effects of virtual scenes. They usually use real-shot panoramas to build scene spaces instead of original model-built scene spaces. However, in When using a panorama, the image resource may contain the hexagonal panorama before stitching, which can be used directly, or there may be only the histogram panorama after stitching, in which case the histogram panorama needs to be re-converted into a hexagonal panorama. However, Most traditional panorama to six-sided panorama conversion can only process one picture at a time. If too many pictures are processed, the processor can only arrange the tasks in order and wait for the previous panorama to be converted before proceeding to the next panorama. conversion, so the traditional conversion method is less efficient and cannot meet actual use needs.

Contents of the invention

The purpose of the present invention is to propose a method, system and storage medium for converting a histogram panorama to a hexagonal image based on OpenCV in order to solve the above-mentioned problem of low efficiency in converting a panorama to a hexagonal image, which can batch process the conversion of panoramic images and make full use of the GPU. Core, the advantage of fast processing speed.

In the first aspect, the present invention achieves the above object through the following technical solution, a method of converting a histogram panorama to a hexahedral image based on OpenCV. The method includes the following steps:

Set the number of converted images and the image resolution, and create a corresponding number of conversion tasks according to the number of images. The conversion task is to realize the conversion of the panorama map to the six-sided map by calling the remap function in OpenCV;

Define a single-sided image in the six-sided image as the target image, obtain the resolution information of the target image, and query whether the resolution information has corresponding mapping data in the local database. If it exists, directly call the Send mapping data to the conversion task;

If it does not exist, calculate new mapping data corresponding to the resolution information. The method of calculating new mapping data corresponding to the resolution information includes:

Obtain the pixel coordinates of the target image and map the pixel coordinates to the plane unit vector [-1, 1] of the Cartesian coordinate system;

Map the coordinates of the plane unit vector [-1, 1] to the spherical unit vector [-√2/2, √2/2];

Use the angular coordinates of the polar coordinate system to map the coordinates of the spherical unit vector [-√2/2, √2/2] to the plane unit vector [-1, 1];

Map from the plane unit vector [-1, 1] to the pixel coordinates of the target image in the texture space as mapping data;

The new mapping data is sent to the conversion task and saved to the local database.

Preferably, the number of images is 6n, where n is the number of panoramas, and each image is numbered while setting the number of images.

Preferably, the conversion task is created by the CPU, and the CPU sends all conversion tasks to different cores of the GPU in sequence for calculation.

Preferably, the method of calculating new mapping data corresponding to the resolution information also includes dividing all pixel coordinates of the target image into two groups of pixel coordinates in the horizontal direction and pixel coordinates in the vertical direction, and simultaneously calculating the corresponding pixel coordinates of each group of pixels. mapping data.

Preferably, the method of calculating new mapping data corresponding to the resolution information sends each calculation to the GPU core for processing by calling an open interface of the GPU.

Preferably, the method also includes pre-saving mapping data corresponding to pictures with resolutions of 1440*1440, 1024*1024, 900*900, 768*768 and 512*512 in a local database.

In the second aspect, the present invention achieves the above object through the following technical solution, a system for converting a histogram panorama into a hexahedral image based on OpenCV. The system includes:

The task creation unit sets the number of converted images and the image resolution, and creates a corresponding number of conversion tasks according to the number of images. The conversion task is to realize the conversion of panoramic map to six-sided map by calling the remap function in OpenCV. ;

A conversion unit used to define a single-sided image among the six-sided images as a target image, obtain the resolution information of the target image, and query whether the resolution information has corresponding mapping data in the local database. If there is , then directly call the mapping data and send it to the conversion task;

If it does not exist, calculate new mapping data corresponding to the resolution information. The method of calculating new mapping data corresponding to the resolution information includes:

Obtain the pixel coordinates of the target image and map the pixel coordinates to the plane unit vector [-1, 1] of the Cartesian coordinate system;

Map the coordinates of the plane unit vector [-1, 1] to the spherical unit vector [-√2/2, √2/2];

Use the angular coordinates of the polar coordinate system to map the coordinates of the spherical unit vector [-√2/2, √2/2] to the plane unit vector [-1, 1];

Map from the plane unit vector [-1, 1] to the pixel coordinates of the target image in the texture space as mapping data;

Send the new mapping data to the conversion task and save it to the local database;

A local database used to save mapping data.

Preferably, the conversion tasks in the conversion unit are created by the CPU, and the CPU sends all conversion tasks to different cores of the GPU in sequence for calculation.

Preferably, the method for calculating new mapping data corresponding to the resolution information in the conversion unit also includes dividing all pixel coordinates of the target image into two groups: pixel coordinates in the horizontal direction and pixel coordinates in the vertical direction, and simultaneously calculating each group. Mapping data corresponding to pixel coordinates.

In a third aspect, the present invention achieves the above object through the following technical solution: a storage medium on which a computer program is stored. When the computer program is executed by a processor, the OpenCV-based histogram as described in the first aspect is implemented. How to convert a panorama to a hexagram.

Compared with the existing technology, the beneficial effects of the present invention are: the present invention calculates the mapping data of each picture of the six-sided picture, and uses the remap function in OpenCV according to the mapping data to realize the rapid conversion of the panorama into the six-sided picture, and in When calculating the mapping data of each picture, first determine whether the mapping data exists in the local database. If it exists, it can be converted by calling it directly. If it does not exist, new mapping data will be calculated. When calculating the mapping data, due to the pixels of the picture, The number of points is huge, so the calculation is put into multiple cores of the GPU and calculated simultaneously. When the number of panoramas is large, the calculation time is greatly reduced. The mapping corresponding to some commonly used resolution pictures can be stored in the local database. Data, using this method can also reduce the resource usage of CPU and GPU.

Description of the drawings

Figure 1 is a flow chart of the method of converting a histogram panorama to a hexagonal image based on OpenCV according to the present invention.

Figure 2 is a schematic diagram of the distribution of model space and panoramic space in the virtual scene of the present invention.

Figure 3 is a schematic diagram of the system for converting a histogram panorama into a hexagonal image based on OpenCV according to the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

Example 1

As shown in Figure 1, a method of converting a histogram panorama to a hexagonal image based on OpenCV includes the following steps:

Step S1: Set the number of converted images and the image resolution, and create a corresponding number of conversion tasks according to the number of images. The conversion task is to convert the panorama into a six-sided map by calling the remap function in OpenCV. The converted image is a single-sided image of the six-sided image. According to the conversion of the panorama to the six-sided image, each panorama can be converted into six images. This step can create a corresponding number of conversion tasks by setting the number of images. The number of images is 6n, where n is the number of panoramas, and each image is numbered while setting the number of images. The purpose of numbering the images is to better sort the converted images, and the numbering can also be used quickly Find the image to prevent the user from being unable to find the corresponding image. The conversion task is a technology that maps the original image to the target image based on the remap function in OpenCV. The remap function is as follows:

void cv::remap(InputArray src,

OutputArray dst,

InputArray map1,

InputArray map2,

int interpolation,

int borderMode = BORDER_CONSTANT,

const Scalar& borderValue = Scalar());

Among them, src: original image; dst: target image, the size is the same as map1, and the data type is the same as src; map1: represents (x, y) coordinate point or x coordinate, type is CV_16SC2, CV_32FC1 or CV_32FC2 ; map2: represents the y coordinate, the type is CV_16UC1, CV_32FC1. When map1 is the (x, y) coordinate, map2 can be empty; Interpolation: represents the interpolation algorithm, the main enumeration type. The INTER_AREA interpolation algorithm is not supported yet. It can be seen from this function that when the original image (histogram panorama), map1 and map2 are input, the target image can be obtained. The histogram panorama is also converted into six views in this way.

Step S2: Define a single-sided image among the six-sided images as the target image, obtain the resolution information of the target image, and query whether the resolution information has corresponding mapping data in the local database. If so, then Proceed directly to step S3;

If it does not exist, calculate new mapping data corresponding to the resolution information. The method of calculating new mapping data corresponding to the resolution information includes:

Step S201: Obtain the pixel coordinates of the target image, and map the pixel coordinates to the plane unit vector [-1, 1] of the Cartesian coordinate system;

Step S202: Map the coordinates of the plane unit vector [-1, 1] to the spherical unit vector [-√2/2, √2/2];

Step S203: Use the angular coordinates of the polar coordinate system to map the coordinates of the spherical unit vector [-√2/2, √2/2] to the plane unit vector [-1, 1];

Step S204: Map from the plane unit vector [-1, 1] to the pixel coordinates of the target image in the texture space as mapping data and save it to the local database;

Step S3: Call the mapping data and send it to the conversion task for execution.

In step S1, according to the remap function, we know that map1 and map2 represent the pixel coordinates of the target image. Therefore, the resolution of the target image is different, and the resolution is used as the pixel point of the image, so the obtained map1 and map2 are also different. In step S2 Map1 and map2 are used as mapping data, so the mapping data of the target image needs to be calculated. There is some mapping data pre-stored in the local database that can be called, so in step S2, the mapping data corresponding to the input target image is now calculated. If the local database search can be found, the mapping data can be directly called and input into the conversion task, which reduces the calculation steps. If it cannot be found, the mapping data can be calculated again. The method also includes pre-saving in the local database mapping data corresponding to pictures with resolutions of 1440*1440, 1024*1024, 900*900, 768*768 and 512*512. Some of the above resolutions are commonly used for image resolution. rate and can be used for most target images. Of course, the resolutions listed above are not all the resolutions pre-saved in the local database. The pre-saved mapping data can be customized according to user habits or needs. I will not list all the resolutions here. Mapping data corresponding to the image resolution.

Figure 2 shows the calculation process of new mapping data. The essence of calculation of mapping data is the conversion of pixel coordinates. Since the target image has different numbers of pixels according to different resolutions, the number of calculation steps required is also huge. This The calculation method can send each calculation to the GPU core for processing by calling the open interface of the GPU, using a large number of GPU cores to perform simultaneous calculations. Even if the number of pixels is huge, the calculation results can be obtained quickly.

Different from the above method of calculating the x and y coordinates of a single pixel, all pixel coordinates of the target image can also be divided into two groups of pixel coordinates in the horizontal direction and pixel coordinates in the vertical direction, and the mapping corresponding to each group of pixel coordinates is calculated at the same time. Data, such as calculating all x coordinates of the target image and calculating all y coordinates of the target image. Although the two calculation methods are different, when input to the conversion task, they are only the difference between map1 and map2, just like map1 described in step S1 : represents (x, y) coordinate point or x coordinate, type is CV_16SC2, CV_32FC1 or CV_32FC2; map2: represents y coordinate, type is CV_16UC1, CV_32FC1. When map1 is (x, y) coordinate, map2 can be empty. In the same way, no matter what kind of calculation method is used, each calculation can be sent to the GPU core for processing by calling the open interface of the GPU to improve calculation efficiency.

The conversion tasks are created by the CPU, and the CPU sends all conversion tasks to different cores of the GPU in sequence for calculation. The conversion tasks can also be accelerated by the GPU to improve conversion efficiency, and the conversion tasks are created and allocated by the CPU, and each task is Independent, it can process multiple panoramas at one time to achieve multi-thread acceleration.

Example 2

As shown in Figure 3, a system based on OpenCV for converting a histogram panorama into a hexagonal image. The system includes:

The task creation unit sets the number of converted images and the image resolution, and creates a corresponding number of conversion tasks according to the number of images. The conversion task is to realize the conversion of panoramic map to six-sided map by calling the remap function in OpenCV. ;The number of images is 6n, where n is the number of panoramas, and each image is numbered while setting the number of images;

A conversion unit used to define a single-sided image among the six-sided images as a target image, obtain the resolution information of the target image, and query whether the resolution information has corresponding mapping data in the local database. If there is , the mapping data is directly called and sent to the conversion task; the mapping data corresponding to pictures with resolutions of 1440*1440, 1024*1024, 900*900, 768*768 and 512*512 is pre-stored in the local database;

If it does not exist, calculate new mapping data corresponding to the resolution information. The method of calculating new mapping data corresponding to the resolution information includes:

Obtain the pixel coordinates of the target image and map the pixel coordinates to the plane unit vector [-1, 1] of the Cartesian coordinate system;

Map the coordinates of the plane unit vector [-1, 1] to the spherical unit vector [-√2/2, √2/2];

Use the angular coordinates of the polar coordinate system to map the coordinates of the spherical unit vector [-√2/2, √2/2] to the plane unit vector [-1, 1];

Map from the plane unit vector [-1, 1] to the pixel coordinates of the target image in the texture space as mapping data;

Send the new mapping data to the conversion task and save it to the local database;

A local database used to save mapping data.

The conversion tasks in the conversion unit are created by the CPU, and the CPU sends all conversion tasks to different cores of the GPU in sequence for calculation.

The method of calculating new mapping data corresponding to the resolution information in the conversion unit also includes dividing all pixel coordinates of the target image into two groups of pixel coordinates in the horizontal direction and pixel coordinates in the vertical direction, and simultaneously calculating the corresponding pixel coordinates of each group of pixels. mapping data. The method of calculating new mapping data corresponding to the resolution information sends each calculation to the GPU core for processing by calling an open interface of the GPU.

Embodiment 2 is essentially the same as Embodiment 1, so the working principles of each unit in Embodiment 2 will not be described in detail.

The method of calculating new mapping data corresponding to the resolution information also includes dividing all pixel coordinates of the target image into two groups of pixel coordinates in the horizontal direction and pixel coordinates in the vertical direction, and simultaneously calculating mapping data corresponding to each group of pixel coordinates. .

Example 3

This embodiment proposes a storage medium. The storage medium includes a storage program area and a storage data area. The storage program area can store operating systems and programs required to run instant messaging functions. The storage data area can store various instant messaging areas. Information and operating instruction sets, etc. The computer program is stored in the stored program area. When the computer program is executed by the processor, the OpenCV-based method of converting a histogram panorama into a hexagonal panorama as described in Embodiment 1 is implemented. The processor may include one or more central processing units, or may be a digital processing unit, or the like.

It is obvious to those skilled in the art that the present invention is not limited to the details of the above-described exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, the embodiments should be regarded as illustrative and non-restrictive from any point of view, and the scope of the present invention is defined by the appended claims rather than the above description, and it is therefore intended that all claims falling within the claims All changes within the meaning and scope of equivalent elements are included in the present invention.

In addition, it should be understood that although this specification is described in terms of implementations, not each implementation only contains an independent technical solution. This description of the specification is only for the sake of clarity, and those skilled in the art should take the specification as a whole. , the technical solutions in each embodiment can also be appropriately combined to form other implementations that can be understood by those skilled in the art.

Claims (10)

1. A method of converting a histogram panorama to a hexagonal image based on OpenCV, which is characterized in that the method includes the following steps: Set the number of converted images and the image resolution, and create a corresponding number of conversion tasks according to the number of images. The conversion task is to realize the conversion of the panorama map to the six-sided map by calling the remap function in OpenCV; Define a single-sided image in the six-sided image as the target image, obtain the resolution information of the target image, and query whether the resolution information has corresponding mapping data in the local database. If it exists, directly call the Send mapping data to the conversion task; If it does not exist, calculate new mapping data corresponding to the resolution information. The method of calculating new mapping data corresponding to the resolution information includes: Obtain the pixel coordinates of the target image and map the pixel coordinates to the plane unit vector [-1, 1] of the Cartesian coordinate system; Map the coordinates of the plane unit vector [-1, 1] to the spherical unit vector [-√2/2, √2/2]; Use the angular coordinates of the polar coordinate system to map the coordinates of the spherical unit vector [-√2/2, √2/2] to the plane unit vector [-1, 1]; Map from the plane unit vector [-1, 1] to the pixel coordinates of the target image in the texture space as mapping data; The new mapping data is sent to the conversion task and saved to the local database. 2. The method of converting a histogram panorama to a hexagonal image based on OpenCV according to claim 1, characterized in that the number of images is 6n, where n is the number of panoramas, and while setting the number of images, each Images are numbered. 3. The method of converting a histogram panorama to a hexagonal image based on OpenCV according to claim 1, characterized in that the conversion task is created by the CPU, and the CPU sends all conversion tasks to different cores of the GPU in sequence for calculation. 4. The method of converting a histogram panorama to a hexagonal panorama based on OpenCV according to claim 1, characterized in that the method of calculating new mapping data corresponding to the resolution information also includes converting all pixel coordinates of the target image It is divided into two groups: horizontal pixel coordinates and vertical pixel coordinates, and the mapping data corresponding to each group of pixel coordinates is calculated at the same time. 5. The method of converting a histogram panorama to a hexagonal image based on OpenCV according to claim 1 or 4, characterized in that the method of calculating the new mapping data corresponding to the resolution information is by calling an open interface of the GPU, Send each calculation to the GPU core for processing. 6. The method of converting a histogram panorama to a hexahedral image based on OpenCV according to claim 1, characterized in that the method also includes pre-saving the resolution in the local database at 1440*1440, 1024*1024, 900*900 , 768*768 and 512*512 images corresponding to the mapping data. 7. A system for converting a histogram panorama into a hexagonal image based on OpenCV, characterized in that the system includes: The task creation unit sets the number of converted images and the image resolution, and creates a corresponding number of conversion tasks according to the number of images. The conversion task is to realize the conversion of panoramic map to six-sided map by calling the remap function in OpenCV. ; A conversion unit used to define a single-sided image among the six-sided images as a target image, obtain the resolution information of the target image, and query whether the resolution information has corresponding mapping data in the local database. If there is , then directly call the mapping data and send it to the conversion task; If it does not exist, calculate new mapping data corresponding to the resolution information. The method of calculating new mapping data corresponding to the resolution information includes: Obtain the pixel coordinates of the target image and map the pixel coordinates to the plane unit vector [-1, 1] of the Cartesian coordinate system; Map the coordinates of the plane unit vector [-1, 1] to the spherical unit vector [-√2/2, √2/2]; Use the angular coordinates of the polar coordinate system to map the coordinates of the spherical unit vector [-√2/2, √2/2] to the plane unit vector [-1, 1]; Map from the plane unit vector [-1, 1] to the pixel coordinates of the target image in the texture space as mapping data; Send the new mapping data to the conversion task and save it to the local database; A local database used to save mapping data. 8. The system for converting a histogram panorama to a hexagonal image based on OpenCV according to claim 7, characterized in that the conversion tasks in the conversion unit are created by the CPU, and the CPU sends all conversion tasks to different cores of the GPU in sequence. Calculation. 9. The system for converting a histogram panorama to a hexagonal panorama based on OpenCV according to claim 7, characterized in that the method for calculating the new mapping data corresponding to the resolution information in the conversion unit also includes converting the target image to All pixel coordinates are divided into two groups: horizontal pixel coordinates and vertical pixel coordinates, and the mapping data corresponding to each group of pixel coordinates is calculated at the same time. 10. A storage medium, characterized in that a computer program is stored thereon, and when the computer program is executed by a processor, the OpenCV-based histogram panorama conversion to six as described in any one of claims 1-6 is implemented. Surface diagram method.