WO2023000221A1

WO2023000221A1 - Free space generation method, movable platform and storage medium

Info

Publication number: WO2023000221A1
Application number: PCT/CN2021/107684
Authority: WO
Inventors: 杨帅; 朱晏辰; 李延召
Original assignee: SZ DJI Technology Co Ltd
Current assignee: SZ DJI Technology Co Ltd
Priority date: 2021-07-21
Filing date: 2021-07-21
Publication date: 2023-01-26
Anticipated expiration: 2024-01-21
Also published as: CN117677976A

Abstract

A free space generation method, comprising: acquiring point cloud data, and generating an initial free space according to the point cloud data (S101); performing convolution on grids in the initial free space, so as to obtain a convolution result (S102); filtering the grids in the initial free space according to the convolution result, so as to obtain candidate free spaces (S103); and determining a plurality of target road edge points in the point cloud data, and generating a target free space according to the plurality of target road edge points and the candidate free spaces (S104). By means of the method, the accuracy of a free space is improved.

Description

Drivable area generation method, movable platform and storage medium

technical field

本申请涉及环境感知领域，尤其涉及一种可行驶区域生成方法、可移动平台及存储介质。The present application relates to the field of environment perception, and in particular to a method for generating a drivable area, a movable platform and a storage medium.

Background technique

可行驶区域(Free Space)是指可移动平台所处环境中可供可移动平台安全探索、到达的区域。尤其是在面对非结构化场景或局部非结构化场景时，可行驶区域可以有效指导可移动平台后续时刻的行动，避免安全事故，保证可移动平台的安全运行。然而，现有的可行驶区域生成方法对可行驶区域的边界噪声滤波效果不好，导致生成的可行驶区域的准确性较低，会影响可移动平台的安全运行，甚至造成安全事故。The free space refers to the area where the mobile platform can safely explore and reach in the environment where the mobile platform is located. Especially in the face of unstructured scenes or partially unstructured scenes, the drivable area can effectively guide the actions of the mobile platform at the next moment, avoid safety accidents, and ensure the safe operation of the mobile platform. However, the existing drivable area generation methods are not effective in filtering the boundary noise of the drivable area, resulting in low accuracy of the generated drivable area, which will affect the safe operation of the mobile platform and even cause safety accidents.

发明内容Contents of the invention

基于此，本申请实施例提供了一种可行驶区域生成方法、可移动平台及存储介质，旨在提高可行驶区域的准确性。Based on this, embodiments of the present application provide a method for generating a drivable area, a movable platform, and a storage medium, aiming at improving the accuracy of the drivable area.

第一方面，本申请实施例提供了一种可行驶区域生成方法，包括：In the first aspect, the embodiment of the present application provides a method for generating a drivable area, including:

获取点云数据，并根据所述点云数据生成初始可行驶区域；Obtaining point cloud data, and generating an initial drivable area according to the point cloud data;

对所述初始可行驶区域中的栅格进行卷积，得到卷积结果；Convolving the grids in the initial drivable area to obtain a convolution result;

根据所述卷积结果，对所述初始可行驶区域中的栅格进行滤波，得到候选可行驶区域；Filtering the grids in the initial drivable area according to the convolution result to obtain a candidate drivable area;

确定所述点云数据中的多个目标路沿点，并根据多个所述目标路沿点和所述候选可行驶区域，生成目标可行驶区域。A plurality of target roadside points in the point cloud data is determined, and a target drivable area is generated according to the plurality of target roadside points and the candidate drivable area.

第二方面，本申请实施例还提供了一种可移动平台，所述可移动平台包括雷达装置、存储器和处理器；In the second aspect, the embodiment of the present application also provides a mobile platform, the mobile platform includes a radar device, a memory, and a processor;

所述雷达装置用于采集点云数据；The radar device is used to collect point cloud data;

所述存储器用于存储计算机程序；The memory is used to store computer programs;

所述处理器，用于执行所述计算机程序并在执行所述计算机程序时，实现以下步骤：The processor is configured to execute the computer program and implement the following steps when executing the computer program:

第三方面，本申请实施例还提供了一种计算机可读存储介质，所述计算机可读存储介质存储有计算机程序，所述计算机程序被处理器执行时使所述处理器实现如上所述的可行驶区域生成方法的步骤。In the third aspect, the embodiment of the present application also provides a computer-readable storage medium, the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor realizes the above-mentioned Steps of a drivable area generation method.

本申请实施例提供了一种可行驶区域生成方法、可移动平台及存储介质，该方法通过对初始可行驶区域中的栅格进行卷积，得到卷积结果，再基于该卷积结果对初始可行驶区域中的栅格进行滤波，可以消除初始可行驶区域内的不合理区域，得到准确的候选可行驶区域，最后确定点云数据中的多个目标路沿点，并基于多个目标路沿点和该候选可行驶区域，生成目标可行驶区域，使得目标可行驶区域内不存在不合理区域，极大地提高了可行驶区域的准确性。The embodiment of the present application provides a method for generating a drivable area, a movable platform, and a storage medium. Filtering the grid in the drivable area can eliminate unreasonable areas in the initial drivable area, obtain accurate candidate drivable areas, and finally determine multiple target roadside points in the point cloud data, and based on multiple target road points A target drivable area is generated along the point and the candidate drivable area, so that there is no unreasonable area in the target drivable area, which greatly improves the accuracy of the drivable area.

应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，并不能限制本申请。It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application.

Description of drawings

为了更清楚地说明本申请实施例技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。In order to illustrate the technical solutions of the embodiments of the present application more clearly, the drawings that need to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present application. Ordinary technicians can also obtain other drawings based on these drawings on the premise of not paying creative work.

图1是实施本申请实施例提供的可行驶区域生成方法的一场景示意图；Fig. 1 is a schematic diagram of a scene implementing the drivable area generation method provided by the embodiment of the present application;

图2是本申请实施例提供的一种可行驶区域生成方法的步骤示意流程图；Fig. 2 is a schematic flowchart of the steps of a method for generating a drivable area provided by an embodiment of the present application;

图3是本申请实施例中的初始可行驶区域的一示意图；Fig. 3 is a schematic diagram of the initial drivable area in the embodiment of the present application;

图4是本申请实施例中的候选可行驶区域的一示意图；Fig. 4 is a schematic diagram of candidate drivable areas in the embodiment of the present application;

图5是本申请实施例中的点云段的一示意图；Fig. 5 is a schematic diagram of the point cloud segment in the embodiment of the present application;

图6是本申请实施例中的点云段的另一示意图；Fig. 6 is another schematic diagram of the point cloud segment in the embodiment of the present application;

图7是本申请实施例中的点云段的另一示意图；Fig. 7 is another schematic diagram of the point cloud segment in the embodiment of the present application;

图8是本申请实施例中的目标可行驶区域的一示意图；Fig. 8 is a schematic diagram of the target drivable area in the embodiment of the present application;

图9是本申请实施例提供的一种可移动平台的结构示意性框图。Fig. 9 is a schematic structural block diagram of a mobile platform provided by an embodiment of the present application.

detailed description

下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of them. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.

附图中所示的流程图仅是示例说明，不是必须包括所有的内容和操作/步骤，也不是必须按所描述的顺序执行。例如，有的操作/步骤还可以分解、组合或部分合并，因此实际执行的顺序有可能根据实际情况改变。The flow charts shown in the drawings are just illustrations, and do not necessarily include all contents and operations/steps, nor must they be performed in the order described. For example, some operations/steps can be decomposed, combined or partly combined, so the actual order of execution may be changed according to the actual situation.

下面结合附图，对本申请的一些实施方式作详细说明。在不冲突的情况下，下述的实施例及实施例中的特征可以相互组合。Some implementations of the present application will be described in detail below in conjunction with the accompanying drawings. In the case of no conflict, the following embodiments and features in the embodiments can be combined with each other.

为解决上述问题，本申请实施例提供了一种可行驶区域生成方法、可移动平台及存储介质，该方法通过对初始可行驶区域中的栅格进行卷积，得到卷积结果，再基于该卷积结果对初始可行驶区域中的栅格进行滤波，可以消除初始可行驶区域内的不合理区域，得到准确的候选可行驶区域，最后确定点云数据中的多个目标路沿点，并基于多个目标路沿点和该候选可行驶区域，生成目标可行驶区域，使得目标可行驶区域内不存在不合理区域，极大地提高了可行驶区域的准确性。In order to solve the above problems, the embodiment of the present application provides a drivable area generation method, a movable platform and a storage medium. The method obtains the convolution result by convolving the grid in the initial drivable area, and then based on the The convolution result filters the grid in the initial drivable area, which can eliminate unreasonable areas in the initial drivable area, obtain accurate candidate drivable areas, and finally determine multiple target roadside points in the point cloud data, and Based on multiple target roadside points and the candidate drivable area, the target drivable area is generated, so that there is no unreasonable area in the target drivable area, and the accuracy of the drivable area is greatly improved.

本申请实施例提供的可行驶区域生成方法可以应用于可移动平台、控制终端、服务器等中，可移动平台可以包括但不限于无人机、可移动机器人和自动驾驶车辆，可移动机器人可以包括扫地机、送餐机器人、无人车等。请参阅图1，图1是实施本申请实施例提供的可行驶区域生成方法的一场景示意图。如图1所示，自动驾驶车辆100包括车辆本体110、动力系统120和雷达装置130，动力系统120和雷达装置130设于车辆本体110上，动力系统120用于为自动驾驶车辆100提供移动动力，雷达装置130用于采集自动驾驶车辆100所处环境的点云数据。The drivable area generation method provided by the embodiment of the present application can be applied to mobile platforms, control terminals, servers, etc. The mobile platforms can include but are not limited to unmanned aerial vehicles, mobile robots and self-driving vehicles, and the mobile robots can include Sweeping machines, food delivery robots, unmanned vehicles, etc. Please refer to FIG. 1 . FIG. 1 is a schematic diagram of a scene implementing the method for generating a drivable area provided by an embodiment of the present application. As shown in FIG. 1 , the self-driving vehicle 100 includes a vehicle body 110 , a power system 120 and a radar device 130 . The power system 120 and the radar device 130 are arranged on the vehicle body 110 . , the radar device 130 is used to collect point cloud data of the environment where the autonomous vehicle 100 is located.

其中，雷达装置130可以包括激光雷达、毫米波雷达。可选的，自动驾驶车辆100可以包括一个或多个雷达装置130。以激光雷达为例，激光雷达可以通过发射激光束探测某个环境中物体的位置、速度等信息，从而获得激光点云。激光雷达可以向包括目标物的环境发射探测信号，然后接受从目标物反射回来的反射信号，根据反射的探测信号、接收到的反射信号，并根据发送和接收的间隔时间等数据参数，获得激光点云。激光点云可以包括N个点，每个点可以包括x,y,z坐标和intensity(反射率)等参数值。Wherein, the radar device 130 may include lidar and millimeter wave radar. Optionally, autonomous vehicle 100 may include one or more radar devices 130 . Taking lidar as an example, lidar can detect the position, speed and other information of objects in an environment by emitting laser beams, so as to obtain laser point clouds. The laser radar can transmit detection signals to the environment including the target object, and then receive the reflected signal reflected from the target object, according to the reflected detection signal, the received reflected signal, and according to the data parameters such as the interval time between sending and receiving, the laser radar can be obtained. point cloud. The laser point cloud can include N points, and each point can include x, y, z coordinates and intensity (reflectivity) and other parameter values.

在一实施例中，自动驾驶车辆100还包括驾驶控制系统(图1未示出)，该驾驶控制系统可以包括一个或多个处理器和传感系统，传感系统用于测量自动驾驶车辆100的位姿信息、运动信息和周围环境信息，一个或多个处理器用于获取点云数据，并根据点云数据生成初始可行驶区域；对初始可行驶区域中的栅格进行卷积，得到卷积结果；根据卷积结果，对初始可行驶区域中的栅格进行滤波，得到候选可行驶区域；确定点云数据中的多个目标路沿点，并根据多个目标路沿点和候选可行驶区域，生成目标可行驶区域。In one embodiment, the self-driving vehicle 100 further includes a driving control system (not shown in FIG. 1 ), which may include one or more processors and a sensing system for measuring One or more processors are used to obtain point cloud data, and generate an initial drivable area based on the point cloud data; convolve the grid in the initial drivable area to obtain the volume According to the convolution result, the grid in the initial drivable area is filtered to obtain the candidate drivable area; multiple target roadside points in the point cloud data are determined, and according to the multiple target roadside points and the candidate drivable area Driving area, generate the target driving area.

以下，将结合图1中的场景对本申请的实施例提供的可行驶区域生成方法进行详细介绍。需知，图1中的场景仅用于解释本申请实施例提供的可行驶区域生成方法，但并不构成对本申请实施例提供的可行驶区域生成方法应用场景的限定。Hereinafter, the method for generating a drivable area provided by the embodiment of the present application will be described in detail in conjunction with the scene in FIG. 1 . It should be noted that the scene in FIG. 1 is only used to explain the drivable area generation method provided by the embodiment of the present application, but does not constitute a limitation on the application scenario of the drivable area generation method provided by the embodiment of the present application.

请参阅图2，图2是本申请实施例提供的一种可行驶区域生成方法的步骤示意流程图。该可行驶区域生成方法可以应用可移动平台，用于生成可行驶区域。Please refer to FIG. 2 . FIG. 2 is a schematic flowchart of steps of a method for generating a drivable area provided by an embodiment of the present application. The method for generating a drivable area can be applied to a movable platform for generating a drivable area.

如图2所示，该可行驶区域生成方法可以包括步骤S101至步骤S104。As shown in FIG. 2 , the method for generating a drivable area may include steps S101 to S104.

步骤S101、获取点云数据，并根据点云数据生成初始可行驶区域。Step S101, acquiring point cloud data, and generating an initial drivable area according to the point cloud data.

其中，获取可移动平台中的雷达装置采集到的点云数据。其中，该雷达装置可以包括激光雷达、毫米波雷达，可移动平台可以包括一个或多个雷达装置。Wherein, the point cloud data collected by the radar device in the movable platform is obtained. Wherein, the radar device may include lidar and millimeter wave radar, and the movable platform may include one or more radar devices.

在一实施例中，从点云数据中提取障碍物点云数据；确定障碍物点云数据中的各障碍物点与可移动平台之间的角度和距离；根据各障碍物点与可移动平台之间的角度和距离，生成初始可行驶区域。In one embodiment, the obstacle point cloud data is extracted from the point cloud data; the angle and distance between each obstacle point in the obstacle point cloud data and the movable platform are determined; The angle and distance between , generating the initial drivable area.

其中，障碍物点云数据中的各障碍物点与可移动平台之间的角度和距离可以基于射线法确定，即确定障碍物点云数据中的可移动平台的当前位置点，以可移动平台的当前位置点为原点，按照预设的角分辨率打出射线，且射线在路径上任一障碍物点云处截止，可以得到射线角度θ与截止距离d的集合为:FS＝{θ _i,d _i}i＝1,2,…,n，然后将射线角度θ与截止距离d的集合投影为栅格图，并将每个方向上在截止距离内的栅格标记为True，得到初始可行驶区域。以水平方向射线为例，角分辨率为1°生成的初始可行驶区域可以如图3所示，图3中多边形的每个顶点代表每条射线的截止点。 Among them, the angle and distance between each obstacle point in the obstacle point cloud data and the movable platform can be determined based on the ray method, that is, the current position point of the movable platform in the obstacle point cloud data is determined, and the movable platform The current position point of is the origin, the ray is shot according to the preset angular resolution, and the ray ends at the point cloud of any obstacle on the path, the set of the ray angle θ and the cut-off distance d can be obtained as: FS={θ _i ,d _i }i=1,2,...,n, then project the set of ray angle θ and cut-off distance d into a grid map, and mark the grids within the cut-off distance in each direction as True, and get the initial drivable area. Taking horizontal rays as an example, the initial drivable area generated with an angular resolution of 1° can be shown in Figure 3. Each vertex of the polygon in Figure 3 represents the cut-off point of each ray.

示例性的，从点云数据中提取障碍物点云数据的方式可以为：将点云数据进行栅格化处理，得到栅格地图，并将栅格地图中的最低点的高度确定为目标高度；将该栅格地图中高度小于或等于目标高度的点确定为候选地面点；基于多个候选地面点进行平面拟合，得到拟合平面，并确定各候选地面点与拟合平面之间的距离；根据各候选地面点与拟合平面之间的距离，从点云数据中提取障碍物点云数据。其中，可以基于平面拟合算法，通过多个候选地面点进行平面拟合，得到拟合平面，平面拟合算法可以包括Ransac算法、最小二乘法。Exemplarily, the method of extracting obstacle point cloud data from point cloud data may be: rasterize the point cloud data to obtain a grid map, and determine the height of the lowest point in the grid map as the target height ; Determine the point whose height is less than or equal to the target height in the grid map as a candidate ground point; carry out plane fitting based on multiple candidate ground points to obtain a fitting plane, and determine the distance between each candidate ground point and the fitting plane Distance: According to the distance between each candidate ground point and the fitting plane, the obstacle point cloud data is extracted from the point cloud data. Wherein, based on a plane fitting algorithm, plane fitting may be performed through multiple candidate ground points to obtain a fitting plane, and the plane fitting algorithm may include a Ransac algorithm and a least square method.

示例性的，根据各候选地面点与拟合平面之间的距离，从点云数据中提取障碍物点云数据的方式可以为：根据各候选地面点与拟合平面之间的距离，在点云数据中标记目标地面点，并在点云数据中剔除各目标地面点，从而可以得到障碍物点云数据。其中，标记的目标地面点包括与拟合平面之间的距离小于或等于预设距离阈值的候选地面点，预设距离阈值可基于实际情况进行设置，本实施例对此不做具体限定。Exemplarily, according to the distance between each candidate ground point and the fitting plane, the method of extracting obstacle point cloud data from the point cloud data can be: according to the distance between each candidate ground point and the fitting plane, at point The target ground points are marked in the cloud data, and each target ground point is eliminated in the point cloud data, so that the obstacle point cloud data can be obtained. Wherein, the marked target ground points include candidate ground points whose distance from the fitting plane is less than or equal to a preset distance threshold, and the preset distance threshold can be set based on actual conditions, which is not specifically limited in this embodiment.

可以理解的是，地面点云与障碍物点云的分割方式除了使用基于平面拟合的方法以外，也可以使用点云的高度差来分割地面点云与障碍物点云，还可以使用基于深度学习的语义分割模型来分割地面点云与障碍物点云，当然，也可以使用其余的方式来分割地面点云与障碍物点云，本申请实施例对此不做具体限定。It can be understood that, in addition to the method based on plane fitting, the segmentation method of the ground point cloud and the obstacle point cloud can also use the height difference of the point cloud to segment the ground point cloud and the obstacle point cloud, and can also use the depth-based The learned semantic segmentation model is used to segment the ground point cloud and the obstacle point cloud. Of course, other methods can also be used to segment the ground point cloud and the obstacle point cloud, which is not specifically limited in this embodiment of the present application.

步骤S102、对初始可行驶区域中的栅格进行卷积，得到卷积结果。Step S102, performing convolution on the grids in the initial drivable area to obtain a convolution result.

示例性的，获取目标卷积算子，其中，目标卷积算子是根据可移动平台的第一尺寸和初始可行驶区域中的栅格的第二尺寸确定的；根据目标卷积算子，对初始可行驶区域中的栅格进行卷积，得到卷积结果。例如，可移动平台的长宽为L和W，初始可行驶区域中的栅格的大小为x，则目标卷积算子的长宽可以表示为：l＝L/x,w＝W/x。需要说明的是，为了减少计算量，可以将目标卷积算子的各边缘位置的元素设置为1，而将除去边缘位置以为的其余位置处的元素设置为0。Exemplarily, a target convolution operator is obtained, wherein the target convolution operator is determined according to the first size of the movable platform and the second size of the grid in the initial drivable area; according to the target convolution operator, Convolve the grid in the initial drivable area to get the convolution result. For example, the length and width of the movable platform are L and W, and the size of the grid in the initial drivable area is x, then the length and width of the target convolution operator can be expressed as: l=L/x, w=W/x . It should be noted that, in order to reduce the amount of calculation, the elements at each edge position of the target convolution operator can be set to 1, and the elements at the remaining positions except the edge positions can be set to 0.

步骤S103、根据卷积结果，对初始可行驶区域中的栅格进行滤波，得到候选可行驶区域。Step S103 , according to the convolution result, filter the grids in the initial drivable area to obtain candidate drivable areas.

示例性的，删除初始可行驶区域中该卷积结果不满足预设条件的栅格，保留该卷积结果满足预设条件的栅格，得到候选可行驶区域。其中，预设条件是根据目标卷积算子中的边缘位置的数量确定的。例如，目标卷积算子包括K个边缘位置，且每个边缘位置的元素均为1，其余位置处的元素均为0，则预设条件包括卷积结果与K的差值小于或等于预设差值阈值。Exemplarily, grids whose convolution results do not meet the preset condition in the initial drivable area are deleted, and grids whose convolution result meets the preset condition are retained to obtain a candidate drivable area. Wherein, the preset condition is determined according to the number of edge positions in the target convolution operator. For example, the target convolution operator includes K edge positions, and the elements at each edge position are 1, and the elements at other positions are all 0, then the preset condition includes that the difference between the convolution result and K is less than or equal to the preset Set the difference threshold.

需要说明的是，通过对初始可行驶区域中的栅格进行卷积，得到卷积结果，再根据卷积结果对初始可行驶区域中的栅格进行滤波，可以消除初始可行驶区域中的不合理栅格。对基于如图3所示的射线图得到的初始可行驶区域进行卷积与滤波后，可以得到如图4所示的候选可行驶区域10。It should be noted that by convoluting the grids in the initial drivable area to obtain the convolution result, and then filtering the grids in the initial drivable area according to the convolution results, the inconsistencies in the initial drivable area can be eliminated. Reasonable grid. After performing convolution and filtering on the initial drivable area obtained based on the ray diagram shown in FIG. 3 , a candidate drivable area 10 as shown in FIG. 4 can be obtained.

步骤S104、确定点云数据中的多个目标路沿点，并根据多个目标路沿点和候选可行驶区域，生成目标可行驶区域。Step S104 , determining multiple target roadside points in the point cloud data, and generating a target drivable area according to the multiple target roadside points and candidate drivable areas.

在一实施例中，沿可移动平台中的雷达装置的扫描路径在点云数据中框选多个连续的点云段；确定点云段的点云距离比例、点云高度差和点云夹角，其中，点云夹角包括经过点云段的中心点的第一侧线段和第二侧线段之间的夹角；根据点云距离比例、点云高度差和点云夹角，从多个点云段中确定多个目标路沿点。In one embodiment, a plurality of continuous point cloud segments are framed in the point cloud data along the scanning path of the radar device in the movable platform; Angle, wherein, the point cloud angle includes the angle between the first side line segment and the second side line segment passing through the center point of the point cloud segment; according to the point cloud distance ratio, point cloud height difference and point cloud angle, from multiple Determine multiple target roadside points in a point cloud segment.

示例性的，沿可移动平台中的雷达装置的扫描路径在点云数据中框选多个连续的点云段的方式可以为：设置多个不同宽度的滑动窗；对于每个滑动窗，沿可移动平台中的雷达装置的扫描路径在点云数据中移动滑动窗，由滑动窗框选点云段，可以得到多个连续的点云段。例如，滑动窗A的宽度为7个点，滑动窗B的宽度为9个点，通过滑动窗A，可以框选得到多个连续的包含7个点的点云段，通过滑动窗B，可以框选得到多个连续的包含9个点的点云段。Exemplarily, the manner of selecting multiple continuous point cloud segments in the point cloud data along the scanning path of the radar device in the movable platform may be as follows: setting a plurality of sliding windows with different widths; for each sliding window, The scanning path of the radar device in the movable platform moves the sliding window in the point cloud data, and the point cloud segment is selected by the sliding window frame to obtain multiple continuous point cloud segments. For example, the width of sliding window A is 7 points, and the width of sliding window B is 9 points. Through sliding window A, multiple continuous point cloud segments containing 7 points can be obtained by frame selection. Through sliding window B, you can Frame selection to obtain multiple continuous point cloud segments containing 9 points.

由于不规则扫描路径的雷达装置扫描路径存在转弯或折返的情况以及雷达装置采集到的点云数据在目标边缘常出现深度差距很大的情况，如果仅使用点云段的点云高度差和点云夹角来确定目标路沿点，会出现较多的错误路沿点，无法保证路沿点的准确性，而通过综合考虑点云段的点云距离比例、点云高度差和点云夹角来确定目标路沿点，可以解决上述问题，从而提高路沿点的准确性，便于后续准确地生成可行驶区域。Due to the fact that the scanning path of the radar device with an irregular scanning path has turning or turning back and the point cloud data collected by the radar device often has a large depth gap at the edge of the target, if only the point cloud height difference and point cloud segment of the point cloud segment are used If the cloud angle is used to determine the target roadside point, there will be more wrong roadside points, and the accuracy of the roadside point cannot be guaranteed. However, by comprehensively considering the point cloud distance ratio, point cloud height difference and point cloud clip The above problem can be solved by using the angle to determine the target roadside point, so as to improve the accuracy of the roadside point and facilitate the subsequent accurate generation of drivable areas.

示例性的，点云高度差包括点云段的中心点与点云段内的其余点之间的最大高度差，第一侧线段经过中心点和点云段内位于中心点的第一方向上的侧点，第二侧线段经过中心点和点云段内位于中心点的第二方向上的侧点。如图5所示，滑动窗20框选的点云段包括7个点，中心点21与点云段内的其余点之间的最大高度差为h，点云夹角为第一侧线段22与第二侧线段23之间的夹角β。Exemplarily, the point cloud height difference includes the maximum height difference between the center point of the point cloud segment and the remaining points in the point cloud segment, the first side line segment passes through the center point and is located in the first direction of the center point in the point cloud segment , the second side line segment passes through the center point and the side point in the second direction of the center point within the point cloud segment. As shown in Figure 5, the point cloud segment selected by the sliding window 20 includes 7 points, the maximum height difference between the center point 21 and the remaining points in the point cloud segment is h, and the angle between the point cloud is the first side line segment 22 The angle β between the line segment 23 and the second side.

示例性的，确定与点云段的中心点相邻的第一侧点以及与中心点相邻的第二侧点，并确定第一侧点和第二侧点之间的第一距离；确定点云段的第一边界点和点云段的第二边界点，并确定第一边界点与第二边界点之间的第二距离；根据第一距离和第二距离，确定点云段的点云距离比例。其中，设第一距离为d ₁，第二距离为d ₂，则点云段的点云距离比例为： Exemplarily, determine the first side point adjacent to the center point of the point cloud segment and the second side point adjacent to the center point, and determine the first distance between the first side point and the second side point; determine The first boundary point of the point cloud segment and the second boundary point of the point cloud segment, and determine the second distance between the first boundary point and the second boundary point; according to the first distance and the second distance, determine the point cloud segment Point cloud distance scale. Among them, assuming that the first distance is d ₁ and the second distance is d ₂ , then the point cloud distance ratio of the point cloud segment is:

示例性的，当点云处于物体边缘时，点云距离比例r会因为点云深度突变而变大，如图6所示，滑动窗30框选的点云段包括7个点，与中心点31的相邻的第一侧点32与第二测点33之间的距离d ₁为第一距离，第一边界点34与第二边界点35之间的距离d ₂为第二距离，因此，点云距离比例r为d ₂/d ₁。 Exemplarily, when the point cloud is at the edge of the object, the point cloud distance ratio r will become larger due to the sudden change of the point cloud depth, as shown in Figure 6, the point cloud segment selected by the sliding window 30 includes 7 points, and the central point The distance d1 between the adjacent first side point 32 of 31 and the second measuring point 33 is the _first distance, and the distance _d2 between the first boundary point 34 and the second boundary point 35 is the second distance, so , the point cloud distance ratio r is d ₂ /d ₁ .

示例性的，在扫描路径折返或转弯处，点云距离比例r将接近于1，且分子分母都较小，如图7所示，滑动窗40框选的点云段包括7个点，与中心点41的相邻的第一侧点42与第二测点43之间的距离d ₁为第一距离，第一边界点44与第二边界点45之间的距离d ₂为第二距离，因此，点云距离比例r为d ₂/d ₁。 Exemplarily, at the turnback or turn of the scanning path, the point cloud distance ratio r will be close to 1, and the numerator and denominator are both smaller, as shown in Figure 7, the point cloud segment selected by the sliding window 40 includes 7 points, and The distance d1 between the adjacent first side point 42 of the central point 41 and the second measuring point 43 is the _first distance, and the distance _d2 between the first boundary point 44 and the second boundary point 45 is the second distance , therefore, the point cloud distance ratio r is d ₂ /d ₁ .

在一实施例中，从多个点云段中选择点云距离比例、点云高度差和点云夹角满足预设路沿点条件的点云段作为目标点云段；根据多个目标点云段确定多个目标路沿点。其中，预设路沿点条件包括点云距离比例位于预设比例范围、点云高度差位于预设高度差范围、点云夹角位于预设夹角范围，预设比例范围、预设高度差范围和预设夹角范围可基于实际情况进行设置，本实施例对此不做具体限定。In one embodiment, select the point cloud segment whose point cloud distance ratio, point cloud height difference and point cloud angle meet the preset roadside point condition from a plurality of point cloud segments as the target point cloud segment; Cloud segments identify multiple target curb points. Among them, the preset roadside point conditions include that the point cloud distance ratio is within the preset ratio range, the point cloud height difference is within the preset height difference range, the point cloud angle is within the preset angle range, the preset ratio range, and the preset height difference The range and the preset included angle range may be set based on actual conditions, which is not specifically limited in this embodiment.

示例性的，根据多个目标点云段确定多个目标路沿点的方式可以为：将各目标点云段内的中心点确定为第一候选路沿点，并确定每个第一候选路沿点与可移动平台之间的角度；根据每个第一候选路沿点与可移动平台之间的角度，将多个第一候选路沿点划分为预设的多个角度区间中的每个角度区间对应的路沿点组；将路沿点组中距离可移动平台最近的候选路沿点确定为目标路沿点，从而可以得到多个目标路沿点。Exemplarily, the manner of determining multiple target roadside points according to multiple target point cloud segments may be as follows: determine the central point in each target point cloud segment as the first candidate roadside point, and determine each first candidate roadside point The angle between the point and the movable platform; according to the angle between each first candidate waypoint and the movable platform, a plurality of first candidate waypoints are divided into each of a plurality of preset angle intervals The curb point group corresponding to each angle interval; the candidate curb point in the curb point group that is closest to the movable platform is determined as the target curb point, so that multiple target curb points can be obtained.

由于路沿一般对称地分布于道路两侧，即在路面没有障碍物的情况下，路沿应该是连接地平面的距离雷达装置(可移动平台)最近的“突起”，因此，通过本实施例的方案可以剔除掉大部分路面以外的误检路沿点，保证路沿点的准确性，便于后续准确地生成可行驶区域。Since the curb is generally symmetrically distributed on both sides of the road, that is, when there is no obstacle on the road surface, the curb should be the "protrusion" closest to the radar device (movable platform) connected to the ground plane, therefore, through this embodiment The scheme can eliminate most of the wrongly detected roadside points other than the road surface, ensure the accuracy of the roadside points, and facilitate the subsequent accurate generation of drivable areas.

其中，预设的多个角度区间可基于实际情况进行设置，本实施例对此不做具体限定。例如，第一候选路沿点分别为路沿点A、路沿点B、路沿点C、路沿点D、路沿点E、路沿点F，预设的角度区间包括角度区间a：0°-120°、角度区间b：121°-240°和角度区间c:241°-360°，路沿点A、路沿点B、路沿点C、路沿点D、路沿点E、路沿点F与可移动平台之间的角度分别为33°、129°、50°、220°、300°、270°，因此，将路沿点A和路沿点C划分为角度区间a：0°-120°所对应的第一路沿点组，将路沿点B和路沿点D划分为角度区间b：121°-240°所对应的第二路沿点组，将路沿点E和路沿点F划分为角度区间c:241°-360°所对应的第三路沿点组。Wherein, the multiple preset angle intervals may be set based on actual conditions, which is not specifically limited in this embodiment. For example, the first candidate curb points are curb point A, curb point B, curb point C, curb point D, curb point E, and curb point F, and the preset angle interval includes angle interval a: 0°-120°, angle interval b: 121°-240° and angle interval c: 241°-360°, curb point A, curb point B, curb point C, curb point D, curb point E , the angles between the roadside point F and the movable platform are 33°, 129°, 50°, 220°, 300°, and 270° respectively, therefore, the roadside point A and the roadside point C are divided into the angle interval a : The first roadside point group corresponding to 0°-120°, divide the roadside point B and the roadside point D into an angle interval b: the second roadside point group corresponding to 121°-240°, divide the roadside point Point E and roadside point F are divided into a third roadside point group corresponding to the angle interval c: 241°-360°.

在一实施例中，将路沿点组中距离可移动平台最近的候选路沿点确定为第二候选路沿点；根据多个第二候选路沿点，更新预设的概率占据栅格地图，其中，该概率占据栅格地图中的栅格占据概率用于表示对应点为路沿点的概率；将更新后的概率占据栅格地图中的栅格占据概率大于预设概率阈值所对应的点确定为目标路沿点。其中，预设概率阈值可基于实际情况进行设置，本实施例对此不做具体限定。通过准确性高的第二候选路沿点来更新概率占据栅格地图，之后再将更新后的概率占据栅格地图中的栅格占据概率大于预设概率阈值所对应的点确定为目标路沿点，可以进一步地提高路沿点的准确性。In one embodiment, the candidate curb point closest to the movable platform in the curb point group is determined as the second candidate curb point; according to a plurality of second candidate curb points, the preset probability occupancy grid map is updated , where the probability of occupying the grid in the grid map is used to indicate the probability that the corresponding point is a roadside point; the updated probability of occupying the grid in the grid map is greater than the preset probability threshold corresponding to The point is determined as the target roadside point. Wherein, the preset probability threshold may be set based on actual conditions, which is not specifically limited in this embodiment. Update the probability occupancy grid map through the second candidate roadside point with high accuracy, and then determine the point corresponding to the grid occupancy probability in the updated probability occupancy grid map greater than the preset probability threshold as the target roadside points, which can further improve the accuracy of roadside points.

示例性的，根据多个第二候选路沿点，更新预设的概率占据栅格地图的方式可以为：确定概率占据栅格地图中存在第二候选路沿点的第一位置以及不存在第二候选路沿点的第二位置；对于第一位置处的点，增大该点的栅格占据概率，对于第二位置处的点，减小该点的栅格占据概率。Exemplarily, according to a plurality of second candidate roadside points, the way of updating the preset probability occupancy grid map may be: determining whether there is a first position of the second candidate roadside point in the probability occupancy grid map and whether there is no second candidate roadside point. The second position of the two candidate roadside points: for the point at the first position, increase the grid occupancy probability of this point, and for the point at the second position, decrease the grid occupancy probability of this point.

在一实施例中，根据多个目标路沿点生成目标路沿轨迹；根据目标路沿轨迹和候选可行驶区域，生成目标可行驶区域。如图8所示，根据多个目标路沿点生成的目标路沿轨迹为路沿轨迹52和路沿轨迹53，且目标可行驶区域51位于路沿轨迹52与路沿轨迹53之间。In an embodiment, a target roadside trajectory is generated according to a plurality of target roadside points; and a target drivable area is generated according to the target roadside trajectory and candidate drivable areas. As shown in FIG. 8 , the target curb trajectory generated according to multiple target curb points is a curb trajectory 52 and a curb trajectory 53 , and the target drivable area 51 is located between the curb trajectory 52 and the curb trajectory 53 .

示例性的，对多个目标路沿点进行曲线拟合，得到候选路沿轨迹，并确定候选路沿轨迹的曲线方程的第一参数；获取历史路沿轨迹的曲线方程的第二参数，其中，历史路沿轨迹是基于上一帧点云数据中的路沿点确定的；根据第二参数和第一参数，确定参数调整值，并根据参数调整值对候选路沿轨迹进行调整，得到目标路沿轨迹。其中，路沿轨迹的拟合算法可以包括RANSAC算法、最小二乘法和霍夫变换算法等。通过综合历史路沿轨迹的曲线方程的参数和候选路沿轨迹的曲线方程的参数来对候选路沿轨迹调整，可以保证路沿轨迹的稳定连续性，提高路沿轨迹的准确性。Exemplarily, curve fitting is performed on multiple target curb points to obtain candidate curb trajectories, and determine the first parameter of the curve equation of the candidate curb trajectories; obtain the second parameter of the curve equation of the historical curb trajectories, wherein , the historical roadside trajectory is determined based on the roadside points in the previous frame of point cloud data; according to the second parameter and the first parameter, determine the parameter adjustment value, and adjust the candidate roadside trajectory according to the parameter adjustment value to obtain the target Curb track. Wherein, the fitting algorithm of the track along the road may include RANSAC algorithm, least square method, Hough transform algorithm and the like. By synthesizing the parameters of the curve equation of the historical curb trajectory and the parameters of the curve equation of the candidate curb trajectory to adjust the candidate curb trajectory, the stability and continuity of the curb trajectory can be ensured and the accuracy of the curb trajectory can be improved.

上述实施例提供的可行驶区域生成方法，通过对初始可行驶区域中的栅格进行卷积，得到卷积结果，再基于该卷积结果对初始可行驶区域中的栅格进行滤波，可以消除初始可行驶区域内的不合理区域，得到准确的候选可行驶区域，最后确定点云数据中的多个目标路沿点，并基于多个目标路沿点和该候选可行驶区域，生成目标可行驶区域，使得目标可行驶区域内不存在不合理区域，极大地提高了可行驶区域的准确性。The drivable area generation method provided by the above embodiment obtains the convolution result by convolving the grids in the initial drivable area, and then filters the grids in the initial drivable area based on the convolution result, which can eliminate The unreasonable areas in the initial drivable area are obtained to obtain accurate candidate drivable areas, and finally multiple target roadside points in the point cloud data are determined, and based on multiple target roadside points and the candidate drivable area, target drivable area is generated. Driving area, so that there is no unreasonable area in the target drivable area, which greatly improves the accuracy of the drivable area.

请参阅图9，图9是本申请实施例提供的一种可移动平台的结构示意性框图。Please refer to FIG. 9 . FIG. 9 is a schematic block diagram of a structure of a mobile platform provided by an embodiment of the present application.

如图9所示，可移动平台300包括雷达装置310、处理器320和存储器330，雷达装置310、处理器320和存储器330通过总线340连接，该总线340比如为I2C(Inter-integrated Circuit)总线。As shown in Figure 9, the movable platform 300 includes a radar device 310, a processor 320 and a memory 330, and the radar device 310, the processor 320 and the memory 330 are connected by a bus 340, such as an I2C (Inter-integrated Circuit) bus .

具体地，雷达装置310可以是激光雷达，也可以是毫米波雷达等，雷达装置310用于采集点云数据。Specifically, the radar device 310 may be a laser radar, or a millimeter-wave radar, etc., and the radar device 310 is used to collect point cloud data.

具体地，处理器320可以是微控制单元(Micro-controller Unit，MCU)、中央处理单元(Central Processing Unit，CPU)或数字信号处理器(Digital Signal Processor，DSP)等。Specifically, the processor 320 may be a micro-controller unit (Micro-controller Unit, MCU), a central processing unit (Central Processing Unit, CPU), or a digital signal processor (Digital Signal Processor, DSP), etc.

具体地，存储器330可以是Flash芯片、只读存储器(ROM，Read-Only Memory)磁盘、光盘、U盘或移动硬盘等。Specifically, the memory 330 may be a Flash chip, a read-only memory (ROM, Read-Only Memory) disk, an optical disk, a U disk, or a mobile hard disk.

其中，所述处理器320用于运行存储在存储器330中的计算机程序，并在执行所述计算机程序时实现以下步骤：Wherein, the processor 320 is configured to run a computer program stored in the memory 330, and implement the following steps when executing the computer program:

可选的，所述处理器在实现对所述初始可行驶区域中的栅格进行卷积，得到卷积结果时，用于实现：Optionally, when the processor performs convolution on the grid in the initial drivable area to obtain a convolution result, it is used to:

获取目标卷积算子，其中，所述目标卷积算子是根据可移动平台的第一尺寸和所述栅格的第二尺寸确定的；acquiring a target convolution operator, wherein the target convolution operator is determined according to the first size of the movable platform and the second size of the grid;

根据所述目标卷积算子，对所述初始可行驶区域中的栅格进行卷积，得到卷积结果。Convolve the grids in the initial drivable area according to the target convolution operator to obtain a convolution result.

可选的，所述处理器在实现根据所述卷积结果，对所述初始可行驶区域中的栅格进行滤波，得到候选可行驶区域时，用于实现：Optionally, when the processor implements filtering the grids in the initial drivable area according to the convolution result to obtain a candidate drivable area, it is used to realize:

删除所述初始可行驶区域中所述卷积结果不满足预设条件的栅格，保留所述卷积结果满足预设条件的栅格，得到候选可行驶区域。Deleting grids whose convolution results do not meet preset conditions in the initial drivable area, and retaining grids whose convolution results meet preset conditions, to obtain candidate drivable areas.

可选的，所述处理器在实现确定所述点云数据中的多个目标路沿点时，用于实现：Optionally, when the processor realizes determining a plurality of target roadside points in the point cloud data, it is used to realize:

沿所述雷达装置的扫描路径在所述点云数据中框选多个连续的点云段；frame selecting a plurality of continuous point cloud segments in the point cloud data along the scanning path of the radar device;

确定所述点云段的点云距离比例、点云高度差和点云夹角，其中，所述点云夹角包括经过所述点云段的中心点的第一侧线段和第二侧线段之间的夹角；Determine the point cloud distance ratio, point cloud height difference and point cloud angle of the point cloud segment, wherein the point cloud angle includes a first side line segment and a second side line segment passing through the center point of the point cloud segment the angle between

根据所述点云距离比例、点云高度差和点云夹角，从多个所述点云段中确定多个目标路沿点。A plurality of target roadside points are determined from the plurality of point cloud segments according to the point cloud distance ratio, the point cloud height difference and the point cloud angle.

可选的，所述处理器在确定所述点云段的点云距离比例时，用于实现：Optionally, when the processor determines the point cloud distance ratio of the point cloud segment, it is used to realize:

确定与所述点云段的中心点相邻的第一侧点以及与所述中心点相邻的第二侧点，并确定所述第一侧点和所述第二侧点之间的第一距离；Determining a first side point adjacent to the center point of the point cloud segment and a second side point adjacent to the center point, and determining a first side point between the first side point and the second side point a distance;

确定所述点云段的第一边界点和所述点云段的第二边界点，并确定所述第一边界点与所述第二边界点之间的第二距离；determining a first boundary point of the point cloud segment and a second boundary point of the point cloud segment, and determining a second distance between the first boundary point and the second boundary point;

根据所述第一距离和所述第二距离，确定所述点云段的点云距离比例。Determine the point cloud distance ratio of the point cloud segment according to the first distance and the second distance.

可选的，所述点云高度差包括所述点云段的中心点与所述点云段内的其余点之间的最大高度差。Optionally, the point cloud height difference includes a maximum height difference between the center point of the point cloud segment and other points in the point cloud segment.

可选的，所述第一侧线段经过所述中心点和所述点云段内位于所述中心点的第一方向上的侧点，所述第二侧线段经过所述中心点和所述点云段内位于所述中心点的第二方向上的侧点。Optionally, the first side line segment passes through the center point and a side point in the point cloud segment located in the first direction of the center point, and the second side line segment passes through the center point and the Side points located in the second direction of the central point within the point cloud segment.

可选的，所述所述处理器在实现根据所述点云距离比例、点云高度差和点云夹角，从多个所述点云段中确定多个目标路沿点时，用于实现：Optionally, when the processor determines a plurality of target roadside points from a plurality of point cloud segments according to the point cloud distance ratio, point cloud height difference and point cloud angle, the processor is used to accomplish:

从多个所述点云段中选择所述点云距离比例、所述点云高度差和所述点云夹角满足预设路沿点条件的点云段作为目标点云段；Select the point cloud segment whose point cloud distance ratio, the point cloud height difference and the point cloud angle satisfy the preset roadside point condition from a plurality of the point cloud segments as the target point cloud segment;

根据多个所述目标点云段确定多个目标路沿点。A plurality of target roadside points are determined according to the plurality of target point cloud segments.

可选的，所述预设路沿点条件包括所述点云距离比例位于预设比例范围、所述点云高度差位于预设高度差范围、所述点云夹角位于预设夹角范围。Optionally, the preset roadside point conditions include that the point cloud distance ratio is in a preset ratio range, the point cloud height difference is in a preset height difference range, and the point cloud angle is in a preset angle range .

可选的，所述处理器在实现根据多个所述目标点云段确定多个目标路沿点时，用于实现：Optionally, when the processor realizes determining a plurality of target roadside points according to a plurality of target point cloud segments, it is used to realize:

将各所述目标点云段内的中心点确定为第一候选路沿点，并确定每个所述第一候选路沿点与所述可移动平台之间的角度；Determining the center point in each of the target point cloud segments as a first candidate roadside point, and determining the angle between each of the first candidate roadside points and the movable platform;

根据所述角度，将多个所述第一候选路沿点划分为预设的多个角度区间中的每个所述角度区间对应的路沿点组；According to the angle, dividing a plurality of the first candidate roadside points into a group of roadside points corresponding to each of the angle intervals in a plurality of preset angle intervals;

将所述路沿点组中距离可移动平台最近的候选路沿点确定为目标路沿点。The candidate wayside point closest to the movable platform in the wayway point group is determined as the target wayside point.

可选的，所述处理器还用于实现以下步骤：Optionally, the processor is also used to implement the following steps:

将所述路沿点组中距离所述可移动平台最近的候选路沿点确定为第二候选路沿点；determining a candidate kerb point closest to the movable platform in the kerb point group as a second candidate kerb point;

根据多个所述第二候选路沿点，更新预设的概率占据栅格地图，其中，所述概率占据栅格地图中的栅格占据概率用于表示对应点为路沿点的概率；Updating a preset probability occupancy grid map according to the plurality of second candidate roadside points, wherein the grid occupancy probability in the probability occupancy grid map is used to represent the probability that the corresponding point is a roadside point;

将更新后的所述概率占据栅格地图中的栅格占据概率大于预设概率阈值所对应的点确定为目标路沿点。A point corresponding to a grid occupancy probability greater than a preset probability threshold in the updated probabilistic occupancy grid map is determined as a target roadside point.

可选的，所述处理器在实现根据多个所述目标路沿点和所述候选可行驶区域，生成目标可行驶区域时，用于实现：Optionally, when the processor realizes generating the target drivable area according to the multiple target roadside points and the candidate drivable areas, it is configured to:

根据多个所述目标路沿点生成目标路沿轨迹；generating a target roadside trajectory according to a plurality of said target roadside points;

根据所述目标路沿轨迹和所述候选可行驶区域，生成目标可行驶区域。A target drivable area is generated according to the target roadside trajectory and the candidate drivable area.

可选的，所述处理器在实现根据多个所述目标路沿点生成目标路沿轨迹时，用于实现：Optionally, when the processor realizes generating the target roadside trajectory according to the multiple target roadside points, it is used to realize:

对多个所述目标路沿点进行曲线拟合，得到候选路沿轨迹，并确定所述候选路沿轨迹的曲线方程的第一参数；Curve fitting is performed on a plurality of said target curb points to obtain candidate curb trajectories, and determine the first parameter of the curve equation of said candidate curb trajectories;

获取历史路沿轨迹的曲线方程的第二参数，其中，所述历史路沿轨迹是基于上一帧点云数据中的路沿点确定的；Obtaining the second parameter of the curve equation of the historical curb trajectory, wherein the historical curb trajectory is determined based on the curb points in the previous frame of point cloud data;

根据所述第二参数和所述第一参数，确定参数调整值，并根据所述参数调整值对所述候选路沿轨迹进行调整，得到目标路沿轨迹。A parameter adjustment value is determined according to the second parameter and the first parameter, and the candidate roadside trajectory is adjusted according to the parameter adjustment value to obtain a target roadside trajectory.

可选的，所述处理器在实现根据所述点云数据生成初始可行驶区域时，用于实现：Optionally, when the processor generates an initial drivable area according to the point cloud data, it is used to:

从所述点云数据中提取障碍物点云数据；Extracting obstacle point cloud data from the point cloud data;

确定所述障碍物点云数据中的各障碍物点与可移动平台之间的角度和距离；Determine the angle and distance between each obstacle point in the obstacle point cloud data and the movable platform;

根据各所述障碍物点与可移动平台之间的角度和距离，生成初始可行驶区域。An initial drivable area is generated according to the angle and distance between each obstacle point and the movable platform.

可选的，所述处理器在实现从所述点云数据中提取障碍物点云数据时，用于实现：Optionally, when the processor realizes extracting obstacle point cloud data from the point cloud data, it is used to realize:

将所述点云数据进行栅格化处理，得到栅格地图，并将所述栅格地图中的最低点的高度确定为目标高度；Rasterizing the point cloud data to obtain a grid map, and determining the height of the lowest point in the grid map as the target height;

将所述栅格地图中高度小于或等于所述目标高度的点确定为候选地面点；Determining a point in the grid map whose height is less than or equal to the target height as a candidate ground point;

基于多个所述候选地面点进行平面拟合，得到拟合平面，并确定各所述候选地面点与所述拟合平面之间的距离；performing plane fitting based on a plurality of candidate ground points to obtain a fitting plane, and determining the distance between each of the candidate ground points and the fitting plane;

根据各所述候选地面点与所述拟合平面之间的距离，从所述点云数据中提取障碍物点云数据。Obstacle point cloud data is extracted from the point cloud data according to the distance between each of the candidate ground points and the fitting plane.

需要说明的是，所属领域的技术人员可以清楚地了解到，为了描述的方便和简洁，上述描述的可移动平台的具体工作过程，可以参考前述可行驶区域生成方法实施例中的对应过程，在此不再赘述。It should be noted that those skilled in the art can clearly understand that for the convenience and brevity of the description, the specific working process of the mobile platform described above can refer to the corresponding process in the aforementioned embodiment of the drivable area generation method. This will not be repeated here.

本申请实施例还提供一种计算机可读存储介质，所述计算机可读存储介质存储有计算机程序，所述计算机程序中包括程序指令，所述处理器执行所述程序指令，实现上述实施例提供的可行驶区域生成方法的步骤。The embodiment of the present application also provides a computer-readable storage medium, the computer-readable storage medium stores a computer program, the computer program includes program instructions, and the processor executes the program instructions to implement the above-mentioned embodiment. The steps of the drivable area generation method of .

其中，所述计算机可读存储介质可以是前述任一实施例所述的可移动平台的内部存储单元，例如所述可移动平台的硬盘或内存。所述计算机可读存储介质也可以是所述可移动平台的外部存储设备，例如所述可移动平台上配备的插接式硬盘，智能存储卡(Smart Media Card，SMC)，安全数字(Secure Digital，SD)卡，闪存卡(Flash Card)等。Wherein, the computer-readable storage medium may be an internal storage unit of the removable platform described in any of the foregoing embodiments, such as a hard disk or a memory of the removable platform. The computer-readable storage medium can also be an external storage device of the removable platform, such as a plug-in hard disk equipped on the removable platform, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital , SD) card, flash memory card (Flash Card), etc.

应当理解，在此本申请说明书中所使用的术语仅仅是出于描述特定实施例的目的而并不意在限制本申请。如在本申请说明书和所附权利要求书中所使用的那样，除非上下文清楚地指明其它情况，否则单数形式的“一”、“一个”及“该”意在包括复数形式。It should be understood that the terms used in the specification of this application are for the purpose of describing specific embodiments only and are not intended to limit the application. As used in this specification and the appended claims, the singular forms "a", "an" and "the" are intended to include plural referents unless the context clearly dictates otherwise.

还应当理解，在本申请说明书和所附权利要求书中使用的术语“和/或”是指相关联列出的项中的一个或多个的任何组合以及所有可能组合，并且包括这些组合。It should also be understood that the term "and/or" used in the description of the present application and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

以上所述，仅为本申请的具体实施方式，但本申请的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到各种等效的修改或替换，这些修改或替换都应涵盖在本申请的保护范围之内。因此，本申请的保护范围应以权利要求的保护范围为准。The above is only a specific embodiment of the application, but the scope of protection of the application is not limited thereto. Any person familiar with the technical field can easily think of various equivalents within the scope of the technology disclosed in the application. Modifications or replacements, these modifications or replacements shall be covered within the scope of protection of this application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

A method for generating a drivable area, comprising:

Obtaining point cloud data, and generating an initial drivable area according to the point cloud data;

Convolving the grids in the initial drivable area to obtain a convolution result;

Filtering the grids in the initial drivable area according to the convolution result to obtain a candidate drivable area;

A plurality of target roadside points in the point cloud data is determined, and a target drivable area is generated according to the plurality of target roadside points and the candidate drivable area.

The method for generating a drivable area according to claim 1, wherein the convolution of the grids in the initial drivable area to obtain a convolution result includes:

acquiring a target convolution operator, wherein the target convolution operator is determined according to the first size of the movable platform and the second size of the grid;

Convolve the grids in the initial drivable area according to the target convolution operator to obtain a convolution result.

The method for generating a drivable area according to claim 1, wherein, according to the convolution result, filtering the grids in the initial drivable area to obtain a candidate drivable area includes:

Deleting grids whose convolution results do not meet preset conditions in the initial drivable area, and retaining grids whose convolution results meet preset conditions, to obtain candidate drivable areas.

The drivable area generation method according to claim 1, wherein said determining a plurality of target roadside points in said point cloud data comprises:

Frame selecting a plurality of continuous point cloud segments in the point cloud data along the scanning path of the radar device in the movable platform;

Determine the point cloud distance ratio, point cloud height difference and point cloud angle of the point cloud segment, wherein the point cloud angle includes a first side line segment and a second side line segment passing through the center point of the point cloud segment the angle between

A plurality of target roadside points are determined from the plurality of point cloud segments according to the point cloud distance ratio, the point cloud height difference and the point cloud angle.

The drivable region generating method according to claim 4, wherein said determining the point cloud distance ratio of said point cloud segment comprises:

Determining a first side point adjacent to the center point of the point cloud segment and a second side point adjacent to the center point, and determining a first side point between the first side point and the second side point a distance;

determining a first boundary point of the point cloud segment and a second boundary point of the point cloud segment, and determining a second distance between the first boundary point and the second boundary point;

Determine the point cloud distance ratio of the point cloud segment according to the first distance and the second distance.

The method for generating a drivable area according to claim 4, wherein the height difference of the point cloud includes a maximum height difference between a central point of the point cloud segment and other points in the point cloud segment.

The method for generating a drivable area according to claim 4, wherein the first side line segment passes through the center point and a side point in the point cloud segment located in the first direction of the center point, so The second side line segment passes through the center point and a side point located in the second direction of the center point in the point cloud segment.

According to the drivable region generation method described in any one of claims 4-7, it is characterized in that, according to the point cloud distance ratio, point cloud height difference and point cloud angle, from a plurality of the point clouds Identify multiple target curb points in the segment, including:

Select the point cloud segment whose point cloud distance ratio, the point cloud height difference and the point cloud angle satisfy the preset roadside point condition from a plurality of the point cloud segments as the target point cloud segment;

A plurality of target roadside points are determined according to the plurality of target point cloud segments.

The drivable area generation method according to claim 8, wherein the preset roadside point conditions include that the point cloud distance ratio is within a preset ratio range, and the point cloud height difference is within a preset height difference range , The included angle of the point cloud is within a preset included angle range.

The drivable region generation method according to claim 8, wherein said determining a plurality of target roadside points according to a plurality of said target point cloud segments comprises:

Determining the center point in each of the target point cloud segments as a first candidate roadside point, and determining the angle between each of the first candidate roadside points and the movable platform;

According to the angle, dividing a plurality of the first candidate roadside points into a group of roadside points corresponding to each of the angle intervals in a plurality of preset angle intervals;

The candidate wayside point closest to the movable platform in the wayway point group is determined as the target wayside point.

The drivable region generating method according to claim 10, wherein the method further comprises:

determining a candidate kerb point closest to the movable platform in the kerb point group as a second candidate kerb point;

Updating a preset probability occupancy grid map according to the plurality of second candidate roadside points, wherein the grid occupancy probability in the probability occupancy grid map is used to represent the probability that the corresponding point is a roadside point;

The point corresponding to the grid occupancy probability in the updated probability occupancy grid map greater than the preset probability threshold is determined as the target roadside point.

The method for generating a drivable area according to any one of claims 1-11, wherein said generating a target drivable area according to a plurality of said target roadside points and said candidate drivable areas includes:

generating a target roadside trajectory according to a plurality of said target roadside points;

A target drivable area is generated according to the target roadside trajectory and the candidate drivable area.

The method for generating a drivable area according to claim 12, wherein said generating a target roadside trajectory according to a plurality of said target roadside points comprises:

Curve fitting is performed on a plurality of said target curb points to obtain candidate curb trajectories, and determine the first parameter of the curve equation of said candidate curb trajectories;

Obtaining the second parameter of the curve equation of the historical curb trajectory, wherein the historical curb trajectory is determined based on the curb points in the previous frame of point cloud data;

A parameter adjustment value is determined according to the second parameter and the first parameter, and the candidate roadside trajectory is adjusted according to the parameter adjustment value to obtain a target roadside trajectory.

The method for generating a drivable area according to any one of claims 1-11, wherein the generating an initial drivable area according to the point cloud data includes:

Extracting obstacle point cloud data from the point cloud data;

Determine the angle and distance between each obstacle point in the obstacle point cloud data and the movable platform;

An initial drivable area is generated according to the angle and distance between each obstacle point and the movable platform.

The drivable region generating method according to claim 14, wherein said extracting obstacle point cloud data from said point cloud data comprises:

Rasterizing the point cloud data to obtain a grid map, and determining the height of the lowest point in the grid map as the target height;

Determining a point in the grid map whose height is less than or equal to the target height as a candidate ground point;

performing plane fitting based on a plurality of candidate ground points to obtain a fitting plane, and determining the distance between each of the candidate ground points and the fitting plane;

Obstacle point cloud data is extracted from the point cloud data according to the distance between each of the candidate ground points and the fitting plane.

A movable platform, characterized in that the movable platform includes a radar device, a memory and a processor;

The radar device is used to collect point cloud data;

The memory is used to store computer programs;

The processor is configured to execute the computer program and implement the following steps when executing the computer program:

The movable platform according to claim 16, characterized in that, when the processor performs convolution on the grid in the initial drivable area to obtain the convolution result, it is used to realize:

The mobile platform according to claim 16, characterized in that, when the processor filters the grids in the initial drivable area according to the convolution result to obtain a candidate drivable area, use To achieve:

The mobile platform according to claim 16, wherein, when the processor realizes determining a plurality of target roadside points in the point cloud data, it is used to realize:

frame selecting a plurality of continuous point cloud segments in the point cloud data along the scanning path of the radar device;

The mobile platform according to claim 19, wherein, when the processor determines the point cloud distance ratio of the point cloud segment, it is used to realize:

The movable platform according to claim 19, wherein the point cloud height difference comprises a maximum height difference between the central point of the point cloud segment and the remaining points in the point cloud segment.

The movable platform according to claim 19, wherein the first side line segment passes through the center point and a side point located in the first direction of the center point in the point cloud segment, and the first side line segment The two side line segments pass through the central point and the side points in the point cloud segment located in the second direction of the central point.

According to the mobile platform described in any one of claims 16-22, it is characterized in that, when the said processor realizes according to the point cloud distance ratio, the point cloud height difference and the point cloud angle, from multiple When multiple target roadside points are determined in the point cloud segment, it is used to realize:

The movable platform according to claim 23, wherein the preset roadside point conditions include that the point cloud distance ratio is within a preset ratio range, the point cloud height difference is within a preset height difference range, and the point cloud height difference is within a preset height difference range. The point cloud included angle is within a preset included angle range.

The mobile platform according to claim 23, characterized in that, when the processor determines a plurality of target roadside points according to a plurality of target point cloud segments, it is used to realize:

The mobile platform according to claim 25, wherein the processor is also used to implement the following steps:

A point corresponding to a grid occupancy probability greater than a preset probability threshold in the updated probabilistic occupancy grid map is determined as a target roadside point.

The mobile platform according to any one of claims 16-26, characterized in that, when the processor generates a target drivable area based on a plurality of target roadside points and the candidate drivable areas , used to implement:

The mobile platform according to claim 27, wherein the processor is configured to realize:

According to the mobile platform according to any one of claims 16-26, it is characterized in that, when the processor realizes generating an initial drivable area according to the point cloud data, it is used to realize:

Extracting obstacle point cloud data from the point cloud data;

The mobile platform according to claim 29, wherein, when the processor realizes extracting obstacle point cloud data from the point cloud data, it is used to realize:

A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor implements any one of claims 1-15. drivable area generation method.