WO2020215172A1

WO2020215172A1 - Obstacle detection method and device, mobile platform, and storage medium

Info

Publication number: WO2020215172A1
Application number: PCT/CN2019/083566
Authority: WO
Inventors: 关雁铭
Original assignee: 深圳市大疆创新科技有限公司
Priority date: 2019-04-20
Filing date: 2019-04-20
Publication date: 2020-10-29
Also published as: CN110799989A

Abstract

Provided in the embodiments of the present invention are an obstacle detection method and device, a mobile platform, and a storage medium, the method comprising: acquiring a first point cloud corresponding to the surrounding environment in which the mobile platform is located; on the basis of size information of the mobile platform, filtering the first point cloud to obtain a second point cloud; projecting the second point cloud onto a two-dimensional plane to obtain at least one projection image; and, on the basis of the at least one projection image, determining obstacle information of the surrounding environment in which the mobile platform is located. By means of the present method, the degree of calculation complexity can be reduced and the efficiency and accuracy of obstacle detection can be increased.

Description

Obstacle detection method, equipment, movable platform and storage medium

Technical field

The invention relates to the field of control technology, in particular to an obstacle detection method, equipment, movable platform and storage medium.

Background technique

At present, with the development of mobile platforms such as unmanned vehicles and sports robots, more and more attention has been paid to the safety of the mobile platforms during the movement. Among them, the detection of obstacles is particularly important. Taking a sports robot as an example, for a sports robot, currently, it is usually possible to detect obstacles on the motion path by means of ray query. This method establishes a ray between the robot and the position to be queried. Starting from the position of the moving robot, the number of points located in a certain radius around the ray is retrieved and counted in the point cloud to determine whether there are obstacles on the path.

However, this type of query method often has the disadvantages of low computational efficiency and low accuracy, resulting in low security of the movable platform during the movement. Therefore, how to better improve the security of the mobile platform is of great significance.

Summary of the invention

The embodiments of the present invention provide an obstacle detection method, equipment, a movable platform, and a storage medium, which can improve obstacle detection efficiency and reduce complexity.

In the first aspect, an embodiment of the present invention provides an obstacle detection method applied to a movable platform, and the method includes:

Acquiring the first point cloud corresponding to the surrounding environment where the movable platform is located;

Filtering the first point cloud to obtain a second point cloud according to the size information of the movable platform;

Project the second point cloud onto a two-dimensional plane to obtain at least one projection image;

According to the at least one projection image, the obstacle information of the surrounding environment where the movable platform is located is determined.

In the second aspect, an embodiment of the present invention provides an obstacle detection device, including a memory and a processor;

The memory is used to store program instructions;

The processor is configured to call the program instructions, and when the program instructions are executed, to perform the following operations:

In a third aspect, an embodiment of the present invention provides a movable platform, and the movable platform includes:

body;

The power system configured on the fuselage is used to provide mobile power for the movable platform;

An obstacle detection device as described in the above second aspect.

In a fourth aspect, an embodiment of the present invention provides an obstacle detection system, including: an obstacle detection device and a movable platform;

The obstacle detection device is used to obtain a first point cloud corresponding to the surrounding environment where the movable platform is located; filter the first point cloud to obtain a second point according to the size information of the movable platform Cloud; project the second point cloud onto a two-dimensional plane to obtain at least one projection image; determine the obstacle information of the surrounding environment where the movable platform is located according to the at least one projection image; and combine the obstacle The information is sent to the mobile platform;

The movable platform is used for moving around obstacles according to the received obstacle information.

In a fifth aspect, an embodiment of the present invention provides a computer-readable storage medium that stores a computer program, and when the computer program is executed by a processor, the method described in the first aspect is implemented.

In the embodiment of the present invention, the obstacle detection device obtains the first point cloud corresponding to the surrounding environment where the movable platform is located, and filters the first point cloud according to the size information of the movable platform to obtain the second point Cloud, reducing the computational complexity; by projecting the second point cloud onto a two-dimensional plane, at least one projection image is obtained, and based on the at least one projection image, the obstacles in the surrounding environment where the movable platform is located are determined Information improves the efficiency and accuracy of obstacle detection.

Description of the drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings in the following description are only some of the present invention. Embodiments, for those of ordinary skill in the art, without creative work, other drawings can be obtained from these drawings.

FIG. 1 is a schematic diagram of a point cloud provided by an embodiment of the present invention;

2 is a schematic side view of a point cloud filter provided by an embodiment of the present invention;

Figure 3a is a schematic plan view of a point cloud provided by an embodiment of the present invention;

Figure 3b is a schematic plan view of a certain area provided by an embodiment of the present invention;

FIG. 3c is a schematic plan view of another determined area provided by an embodiment of the present invention;

Figure 4a is a schematic diagram of vote counting provided by an embodiment of the present invention;

Figure 4b is a schematic diagram of a region division provided by an embodiment of the present invention;

Figure 5 is a schematic structural diagram of an obstacle detection system provided by an embodiment of the present invention;

6 is a schematic flowchart of an obstacle detection method provided by an embodiment of the present invention;

Fig. 7 is a schematic structural diagram of an obstacle detection device provided by an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention.

Hereinafter, some embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the case of no conflict, the following embodiments and features in the embodiments can be combined with each other.

The obstacle detection method provided in the embodiment of the present invention may be executed by an obstacle detection system, and specifically, may be executed by an obstacle detection device in the obstacle detection system. Wherein, the obstacle detection system includes an obstacle detection device and a movable platform. In some embodiments, the obstacle detection device may be installed on a movable platform; in some embodiments, the obstacle detection device may be spatially independent of the movable platform; in some embodiments The obstacle detection device may be a component of a movable platform, that is, the movable platform includes an obstacle detection device.

In other embodiments, the obstacle detection method can also be applied to other mobile devices, such as mobile devices that can move autonomously, such as robots, unmanned vehicles, and unmanned ships.

The obstacle detection device in the obstacle detection system can obtain the first point cloud corresponding to the surrounding environment where the movable platform is located, as shown in FIG. 1, which is a schematic diagram of a point cloud provided by an embodiment of the present invention, wherein , The point cloud in FIG. 1 is the acquired first point cloud corresponding to the surrounding environment where the movable platform 11 is located. In some embodiments, the first point cloud may be obtained through lidar, or may be obtained through a camera on a movable platform, which is not specifically limited in the embodiment of the present invention.

In the embodiment of the present invention, considering the plane constraint of the movable platform on the plane, the acquired point cloud contains a considerable amount of redundant information. Therefore, after acquiring the first point cloud, the obstacle detection device The first point cloud may be pre-processed. When the obstacle detection device preprocesses the first point cloud, it can obtain the size information of the movable platform, and filter the first point cloud according to the size information of the movable platform to obtain The second point cloud.

In one embodiment, the size information of the movable platform includes the height of the movable platform, and the obstacle detection device filters the first point cloud to obtain the first point cloud according to the size information of the movable platform. In the case of two point clouds, the height of the movable platform can be obtained, and according to the height of the movable platform, the area above the height of the movable platform is determined as the deletion area, and the point cloud in the deletion area is deleted. Delete to obtain the second point cloud.

In one embodiment, the size information of the movable platform includes the safe crossing height of the movable platform, and the obstacle detection device is filtering the first point cloud according to the size information of the movable platform. When the second point cloud is obtained, the safe crossing height of the movable platform can be obtained, and according to the safe crossing height of the movable platform, the area below the safe crossing height of the movable platform is determined as the deletion area, and Delete the point cloud in the deletion area to obtain the second point cloud.

In one embodiment, the size information of the movable platform includes the height of the movable platform and the safe crossing height, and the obstacle detection device performs the measurement on the first point cloud according to the size information of the movable platform. When filtering to obtain the second point cloud, the height of the movable platform and the safe crossing height can be obtained, and the area above the height of the movable platform is determined as the first deletion area, and the area below the safe crossing height It is the second deletion area, and the point clouds in the first deletion area and the second deletion area are deleted to obtain the second point cloud.

Specifically, Figure 2 is used as an example to illustrate. Figure 2 is a schematic side view of a point cloud filter provided by an embodiment of the present invention. As shown in Figure 2, according to the safe crossing height position 22 and the height position 23 of the movable platform 20, Determine the first deletion area 24 and the second deletion area 25 outside the area between the safe crossing height position 22 and the height position 23 of the movable platform, and place the first deletion area 24 and the second deletion area 25 Delete the point cloud to obtain the second point cloud area 26 between the safe crossing height position 22 and the height position 23 of the movable platform, and determine that the point cloud in the second point cloud area 26 is The second point cloud.

Optionally, taking an unmanned vehicle as an example, the safe crossing height may be the chassis height of the unmanned vehicle, or a height determined based on the chassis height of the unmanned vehicle.

For planar motion robots, obstacles above the height position 23 of the movable platform will not affect the movement of the movable platform; obstacles between the ground position 21 and the safe crossing height position 22 can be passed by the movable platform smoothly. The point cloud corresponding to the obstacle is a redundant point cloud. Deleting this part of the point cloud can reduce the complexity of calculating the point cloud, thereby improving the efficiency of obstacle detection. It can be seen that by filtering the acquired first point cloud of the surrounding environment of the movable platform, redundant point clouds can be filtered out, so as to reduce the complexity of calculating point clouds during subsequent obstacle detection and improve obstacle detection. accuracy.

In the embodiment of the present invention, after the obstacle detection device obtains the second point cloud through filtering, it may project the second point cloud onto a two-dimensional plane to obtain at least one projection image. It can be specifically described with reference to Figure 2 and Figure 3a. Figure 3a is a schematic plan view of a point cloud provided by an embodiment of the present invention. The obstacle detection device obtains the information in the second point cloud area 26 of Figure 2 through filtering. After the second point cloud, the second point cloud in the second point cloud area 26 can be projected onto a two-dimensional plane as shown in FIG. 3a to obtain a projected image 31.

In the embodiment of the present invention, the obstacle detection device may determine the obstacle information of the surrounding environment where the movable platform is located according to the at least one projection image.

In one embodiment, when the obstacle detection device determines the obstacle information of the surrounding environment in which the movable platform is located according to the projection image, it may obtain information from the projection image according to the type of the movable platform. Determine the appropriate area in the, and divide the area into multiple grid areas. The obstacle detection device may determine whether each grid area is an obstacle area according to the number of point clouds and/or depth information in each grid area. In some embodiments, the depth information is the distance from the second point cloud to the movable platform.

In some embodiments, the types of the movable platform include, but are not limited to, omnidirectional motion robots, non-omnidirectional motion robots (such as three-wheel mobile robots), and the like. If the type of the movable platform is an omnidirectional motion robot, the obstacle detection device may use the position of the omnidirectional motion robot as the geometric center to determine the area. If the type of the movable platform is a non-omnidirectional motion robot, the obstacle detection device may determine the area using the position of the non-omnidirectional motion robot as a bottom boundary point.

Specifically, Figure 3b is an example. Figure 3b is a schematic plan view of a certain area provided by an embodiment of the present invention. Assuming that the type of the movable platform is an omnidirectional motion robot, if the obstacle detection device obtains the omnidirectional motion robot, If the current position 32 of the omnidirectional movement robot is the geometric center, the current position 32 of the omnidirectional movement robot can be used as the geometric center to determine the area 33 from the projection image 31 and divide the area 33 into multiple grid areas.

Taking Figure 3c as an example again, Figure 3c is another schematic plan view of a certain area provided by an embodiment of the present invention. Assuming that the type of the movable platform is a non-omnidirectional motion robot, if the obstacle detection device acquires all According to the current position 34 of the non-omnidirectional motion robot, the current position 34 of the non-omnidirectional motion robot can be the bottom boundary point of the region, and the region 35 is determined from the projection image 31, and the region 35 is divided into multiple Grid areas.

In an embodiment, when the obstacle detection device divides the area into a plurality of grid areas, the division may be performed according to the size information of the movable platform. The grid area smaller than the size of the movable platform does not have much value, so the grid area can be divided according to the size information larger than the movable platform.

In one embodiment, if there is a point cloud in the grid area, it means that there is an obstacle, but actually due to noise and misdetection, a considerable part of the point cloud in the grid area does not represent the actual obstacle.

In one embodiment, when the point cloud is acquired through the camera, due to the characteristics of objects in the image that are close and far small, for obstacle A and obstacle B of the same size, the number of point clouds corresponding to obstacle A in the distance is small , The nearby obstacle B corresponds to a large number of point clouds. If the number of point clouds is used as the obstacle judgment criterion, it may cause the distant obstacle A to be detected by mistake. Therefore, in the embodiment of the present invention, the obstacle detection device can determine the voting information of each point cloud in each grid area according to the depth information of the point cloud, and according to the number of the point clouds and the voting information Information, the evaluation parameter of each grid area is determined, and the evaluation parameter is compared with a preset parameter to determine that a grid area with the evaluation parameter greater than the preset parameter is the obstacle area. In some embodiments, the evaluation parameters may include, but are not limited to, numbers, percentages, etc. determined based on voting information.

In some embodiments, the obstacle detection device may determine areas such as free areas and unknown areas in the grid area according to the evaluation parameters. The embodiment of the present invention does not specifically limit the division of areas. In some embodiments, the unknown area may be a grid area without a second point cloud; in some embodiments, the free area may be a grid area where the evaluation parameter is smaller than the preset parameter .

It can be specifically described with reference to FIGS. 4a and 4b. FIG. 4a is a schematic diagram of a vote counting provided by an embodiment of the present invention, and FIG. 4b is a schematic diagram of a region division provided by an embodiment of the present invention. As shown in Figure 4a, the voting information of the grid area 411, the grid area 412, the grid area 413, and the grid area 414 are all 6, that is, the number of votes for each point cloud in each grid is 6 votes. , There are a total of 5 point clouds in the grid area 411, and a total of 30 votes will be cast for the 5 point clouds. The evaluation parameter of the grid area 411 is 30; there are 6 point clouds in the grid area 412, and the 6 point clouds total 36 votes were cast, and the evaluation parameter of the grid area 412 was 36; there were 8 point clouds in the grid area 413, and a total of 48 votes were cast for the 6 point clouds, and the evaluation parameter of the grid area 413 was 48; There are 9 point clouds in 414, 54 votes are cast for 9 point clouds, and the evaluation parameter of the grid area 414 is 54. The voting information of the grid area 415 is 5, that is, the number of votes for each point cloud in the grid area 415 is 5 votes. There is 1 point cloud in the grid area 415, and a total of 5 votes are cast for 1 point cloud. , The evaluation parameter of the grid area 415 is 5. If the preset parameter is 25, it can be determined that the grid area 411, the grid area 412, the grid area 413, and the grid area 414 are all obstacle areas, and the grid area 415 is a free area. Thus, the obstacle area 41 (black marked area), the free area 42 (light gray marked area), and the unknown area 43 (dark gray marked area) as shown in FIG. 4b can be determined.

In the embodiment of the present invention, by filtering the acquired first point cloud corresponding to the surrounding environment where the movable platform is located according to the size information of the movable platform to obtain the second point cloud, redundant point clouds can be filtered out. Improve the efficiency of point cloud processing and reduce the computational complexity; by projecting the second point cloud onto a two-dimensional plane to obtain at least one projection image, the distribution of the point cloud can be viewed from the two-dimensional plane more intuitively and conveniently; The at least one projection image determines a plurality of grid areas, and further determines obstacle information according to the number of point clouds and/or depth information of each grid area, which can improve the efficiency and accuracy of obstacle detection.

The obstacle detection system provided by the embodiment of the present invention will be schematically described below with reference to FIG. 5.

Please refer to FIG. 5, which is a schematic structural diagram of an obstacle detection system provided by an embodiment of the present invention. The obstacle detection system includes: an obstacle detection device 51 and a movable platform 52. Wherein, a communication connection between the movable platform 52 and the obstacle detection device 51 can be established through a wireless communication connection. Wherein, in some scenarios, a communication connection between the movable platform 52 and the obstacle detection device 51 may also be established through a wired communication connection. The movable platform 52 may be a movable device such as an unmanned vehicle, an unmanned ship, and a movable robot. The movable platform 52 includes a power system 521, and the power system 521 is used to provide the movable platform 52 with moving power. In other embodiments, the movable platform 52 and the obstacle detection device 51 are independent of each other. For example, the obstacle detection device 51 is set in a cloud server and establishes a communication connection with the movable platform 52 through a wireless communication connection.

In the embodiment of the present invention, the obstacle detection device 51 can obtain the first point cloud corresponding to the surrounding environment where the movable platform 52 is located; according to the size information of the movable platform 52, the first point cloud Filtering to obtain a second point cloud; projecting the second point cloud onto a two-dimensional plane to obtain at least one projection image; according to the at least one projection image, determining obstacles in the surrounding environment where the movable platform 52 is located Information to improve the efficiency and accuracy of obstacle detection, and improve the safety of the movable platform 52 during the movement.

The obstacle detection method provided by the embodiment of the present invention will be schematically described below with reference to FIG. 6 and FIG. 7.

Please refer to FIG. 6 for details. FIG. 6 is a schematic flowchart of an obstacle detection method according to an embodiment of the present invention. The method may be executed by an obstacle detection device, and the specific explanation of the obstacle detection device is as described above. Specifically, the method of the embodiment of the present invention includes the following steps.

S601: Acquire a first point cloud corresponding to the surrounding environment where the movable platform is located.

In the embodiment of the present invention, the obstacle detection device can obtain the first point cloud corresponding to the surrounding environment where the movable platform is located. Taking FIG. 1 as an example, the point cloud shown in FIG. 1 is the acquired first point cloud of the surrounding environment where the movable platform 11 is located.

In one embodiment, when the obstacle detection device obtains the first point cloud corresponding to the surrounding environment where the movable platform is located, it may obtain the first point corresponding to the surrounding environment where the movable platform is located through lidar cloud.

In some embodiments, the lidar is a perceptual sensor that can obtain three-dimensional information of the scene. The basic principle is to actively emit laser pulse signals to the detected object and obtain the reflected pulse signals. According to the time difference between the transmitted signal and the received signal, the depth information of the distance detector of the object to be measured is calculated; Know the launch direction, obtain the angle information of the measured object relative to the lidar; combine the aforementioned depth information and angle information to obtain a large number of detection points (called point clouds), based on the point cloud, the spatial three-dimensional information of the measured object relative to the lidar can be reconstructed .

In one embodiment, when the obstacle detection device obtains the first point cloud corresponding to the surrounding environment where the movable platform is located, it may obtain the first point cloud corresponding to the surrounding environment where the movable platform is located through a camera. . In some embodiments, the camera may be mounted on the movable platform. In some embodiments, the camera may also be independent of the movable platform and installed in the environment where the movable platform is located. In some embodiments, the camera includes, but is not limited to, binocular cameras, monocular cameras, TOF cameras and other camera devices.

In some embodiments, when the obstacle detection device acquires the first point cloud corresponding to the surrounding environment where the movable platform is located through the camera, it may convert the point cloud acquired by the camera to the world based on a preset conversion matrix. In the coordinate system, the first point cloud corresponding to the surrounding environment of the movable platform is obtained; wherein, the preset conversion matrix includes an internal parameter matrix and an external parameter matrix, and the external parameter matrix includes a rotation matrix and/or a translation vector . In some embodiments, when the origin of the world coordinate system is set on the movable platform, the external parameter matrix only includes a rotation matrix.

In some embodiments, the internal parameter matrix is determined based on a plurality of internal parameters, and the internal parameters are parameters obtained by camera calibration, such as focal length and principal point coordinates. In some embodiments, the external parameter matrix may include a rotation matrix and/or a translation vector, wherein the rotation matrix may be determined by the posture of the camera, and the translation vector may be determined by the positioning information of the camera.

It can be seen that the embodiment of the present invention converts the point cloud collected by the camera to the world coordinate system to obtain the first point cloud. In the process of converting the point cloud collected by the camera to the world coordinate system, the camera can collect The obtained point cloud undergoes processing such as distortion removal, thereby improving the accuracy of the first point cloud.

S602: Filter the first point cloud according to the size information of the movable platform to obtain a second point cloud.

In the embodiment of the present invention, the obstacle detection device may filter the first point cloud to obtain the second point cloud according to the size information of the movable platform.

In one embodiment, the size information includes the height of the movable platform, and the obstacle detection device filters the first point cloud to obtain the second point according to the size information of the movable platform. In case of cloud, the first point cloud may be filtered according to the height of the movable platform to obtain the second point cloud.

In one embodiment, the size information of the movable platform includes the safe crossing height of the movable platform, and the obstacle detection device is filtering the first point cloud according to the size information of the movable platform. When the second point cloud is obtained, the first point cloud may be filtered according to the safe crossing height of the movable platform to obtain the second point cloud.

In one embodiment, the size information of the movable platform includes the height of the movable platform and the height of safe crossing, and the obstacle detection device performs the measurement on the first point cloud according to the size information of the movable platform. When filtering to obtain the second point cloud, the first point cloud may be filtered according to the height of the movable platform and the safe crossing height to obtain the second point cloud.

Taking FIG. 2 as an example, according to the safe crossing height position 22 and the height position 23 of the movable platform 20, the first deletion outside the area between the safe crossing height position 22 and the height position 23 of the movable platform can be determined Area 24 and the second deletion area 25, and delete the point cloud in the first deletion area 24 and the second deletion area 25 to obtain the distance between the safe crossing height position 22 and the height position 23 of the movable platform And determine that the point cloud in the second point cloud area 26 is the second point cloud. It can be seen that in the embodiment of the present invention, the first point cloud is filtered by the size information of the movable platform, redundant point clouds can be deleted, and the subsequent calculation of the number of point clouds and the complexity of depth information are reduced.

S603: Project the second point cloud onto a two-dimensional plane to obtain at least one projection image.

In the embodiment of the present invention, the obstacle detection device may project the second point cloud onto a two-dimensional plane to obtain at least one projection image.

In an embodiment, the two-dimensional plane may include a horizontal plane, and the obstacle detection device may project the second point cloud onto the horizontal plane to obtain at least one projection image. In some embodiments, the horizontal plane is a plane parallel to the position where the movable platform is located. Taking an unmanned vehicle as an example, assuming that the drone is currently climbing a slope, the horizontal plane is a plane parallel to the ramp where the unmanned vehicle is currently located.

In one embodiment, the two-dimensional plane may include the ground, and the obstacle detection device may project the second point cloud onto the ground to obtain at least one projected image. In some embodiments, the ground is The ground parallel to the movable platform. Taking an unmanned vehicle as an example, when the unmanned vehicle is driving on a level ground, the two-dimensional plane may be the ground.

It can be seen that, by projecting the second point cloud in the world coordinate system to a two-dimensional plane, the embodiment of the present invention satisfies the plane mobility of the movable platform, and facilitates intuitive calculation of the point cloud on the plane.

S604: Determine obstacle information of the surrounding environment where the movable platform is located according to the at least one projection image.

In one embodiment, when the obstacle detection device determines the obstacle information of the surrounding environment where the movable platform is located according to the projection image, the projection image may be divided into multiple grid areas, and According to the point cloud in each grid area, it is determined whether each grid area is an obstacle area. The specific implementation is as described above and will not be repeated here.

In one embodiment, when the obstacle detection device divides the projection image into multiple grid areas, the type of the movable platform may be acquired, and the type of the movable platform may be determined according to the type of the movable platform. The location of each grid area, the specific implementation is as described above, and will not be repeated here.

In one embodiment, the type of the movable platform includes an omnidirectional motion robot, and the obstacle detection device can obtain all the positions of the multiple grid areas according to the type of the movable platform. The current position of the omnidirectional motion robot, and the current position of the omnidirectional motion robot is used as a geometric center to determine the positions of the multiple grid regions.

Specifically, Figure 3b can be taken as an example. Assuming that the movable platform is an omnidirectional motion robot, if the current position 32 of the omnidirectional motion robot, the obstacle detection device can determine that the current position 32 of the omnidirectional motion robot is geometric In the center, an area 33 is determined from the projected image 31, and the area 33 is divided into multiple grid areas.

In one embodiment, the type of the movable platform includes a non-omnidirectional motion robot, and the obstacle detection device can obtain the position of the multiple grid areas according to the type of the movable platform. The current position of the non-omnidirectional motion robot is determined by using the current position of the non-omnidirectional motion robot as a boundary point to determine the positions of the multiple grid regions. In some embodiments, the boundary point may be a boundary point of any side of the determined area, and the specific boundary point of which side is determined according to the type of the non-omnidirectional motion robot.

Specifically, Figure 3c can be taken as an example. Assuming that the movable platform is a non-omnidirectional motion robot, and the non-omnidirectional motion robot is a three-wheeled mobile robot, if the current position of the three-wheeled mobile robot is 34, obstacle detection The device may determine that the current position 34 of the three-wheeled mobile robot is the boundary point in the bottom boundary line, and determine from the projection image 31 that the current position 34 of the three-wheeled mobile robot is the boundary in the bottom boundary line Point the area 33 and divide the area 33 into multiple grid areas.

It can be seen that the embodiment of the present invention determines the positions of the multiple grid areas according to the type of the movable platform, and can determine the grid area according to the motion characteristics of different types of movable platforms to ensure different types of movable platforms. The main moving area of is within the multiple grid areas, so as to avoid that the moving area of the movable platform is outside the multiple grid areas and hits an undetected obstacle outside the multiple grid areas, thereby Improve the safety of different types of movable platforms during the movement.

In one embodiment, when the obstacle detection device divides the projection image into multiple grid areas, it can obtain the size of the movable platform, and determine the size of the movable platform according to the size of the movable platform. The size of each grid area in multiple grid areas. The division of the grid area smaller than the size of the movable platform does not have much practical value. Therefore, the embodiment of the present invention divides the multiple grids in such a way that the size of each grid area is larger than the size of the movable platform. Grid area. The specific implementation is as described above and will not be repeated here.

It can be seen that the embodiment of the present invention divides multiple grid areas according to the size of each grid area larger than the size of the movable platform, which can further reduce the calculation amount required for obstacle detection and improve the efficiency of obstacle detection.

In one embodiment, when the obstacle detection device determines whether each grid area is an obstacle area according to the point cloud in each grid area, it can acquire the midpoint of each grid area. The number of clouds. When the number of point clouds is greater than a preset number threshold, it can be determined that the grid area is an obstacle area.

Taking Figure 4a as an example, assuming that the preset number threshold is 10, if the number of point clouds in the grid area 411 is 15, the number of point clouds in the grid area 412 is 16, the number of point clouds in the grid area 413 is 18, If the number of point clouds in the area 414 is 20, it can be determined that the number of point clouds in the grid area 411, the grid area 412, the grid area 413, and the grid area 414 are all greater than the preset number threshold of 10. The area composed of the grid area 411, the grid area 412, the grid area 413, and the grid area 414 is an obstacle area.

It can be seen that the embodiment of the present invention determines whether the grid area is an obstacle area by judging whether the point cloud data in each grid area is greater than a preset number threshold, ensuring that the grid area is determined only when there are enough point clouds. It is an obstacle area, which improves the accuracy of obstacle detection.

In one embodiment, when the obstacle detection device determines whether each grid area is an obstacle area according to the point cloud in each grid area, it can acquire the midpoint of each grid area. According to the number and depth information of the cloud, it is determined whether the grid area is an obstacle area according to the number and depth information of the point cloud.

In one embodiment, when the obstacle detection device determines whether the grid area is an obstacle area according to the number and depth information of the point cloud, it may determine the depth information of the point cloud. The voting information of each point cloud in each grid area, and the evaluation parameters of each grid area are determined according to the number of the point clouds and the voting information, and the evaluation parameters are combined with preset parameters The comparison is performed to determine that the grid area with the evaluation parameter greater than the preset parameter is the obstacle area. In some embodiments, the obstacle may include, but is not limited to, any one or more objects that hinder the movement of the movable platform, such as fixed buildings, other movable equipment, and ground facilities. In some embodiments, the grid area may include, but is not limited to, any one or more of obstacle areas, free areas, and unknown areas.

In one embodiment, the obstacle detection device can obtain the depth information of each point cloud in each grid area, and calculate the average depth of each point cloud in each grid area, and according to the preset depth According to the correspondence relationship with the number of votes, the number of votes corresponding to the depth average value is determined, so that the total number of votes obtained by all point clouds in each grid area is determined according to the number of votes corresponding to the depth average value. The obstacle detection device may determine the evaluation parameter corresponding to the total number of votes according to the corresponding relationship between the preset total number of votes and the evaluation parameters, and compare the evaluation parameters with the preset parameters. By setting the parameters, the grid area can be determined as an obstacle area.

Figure 4a can be taken as an example. Assuming that there are some clouds A, B, and C in the grid area 411, and the depth is 1.5m, 1.6m, 1.7m, the obstacle detection device can detect the point clouds A, B in the grid area 411. , The depth of C is averaged, and the calculated depth average is 1.6m. If the average depth of 1.6m can correspond to a number of votes such as 2 votes, then the three point clouds A, B, and C in the grid area 411 total 6 votes are cast. If the evaluation parameter corresponding to the 6 votes is 6 and the preset parameter is 5, it can be determined that the grid area 411 is an obstacle area.

In one embodiment, the obstacle detection device can obtain the depth information of each point cloud in each grid area, and determine the depth information of each point cloud according to the preset corresponding relationship between the depth and the number of votes. According to the corresponding number of votes, the total number of votes obtained by all point clouds in each grid area is determined according to the number of votes corresponding to the depth information of each point cloud. The obstacle detection device may determine the evaluation parameter corresponding to the total number of votes according to the corresponding relationship between the preset total number of votes and the evaluation parameters, and compare the evaluation parameters with the preset parameters. Set the parameters, you can determine that the grid area is an obstacle area.

Taking Figure 4a as an example, suppose there are some clouds A, B, and C in the grid area 411 with depths of 1.5m, 1.6m, and 1.7m. If the number of votes corresponding to the depth of 1.5m is 1 vote, the depth of 1.6m The corresponding number of votes is 2 votes, and the number of votes corresponding to the depth of 1.7m is 3 votes. It can be determined that the three point clouds A, B, and C in the grid area 411 cast a total of 6 votes. If 6 votes The corresponding evaluation parameter is 6, and the preset parameter is 5. It can be determined that the grid area 411 is an obstacle area.

In some embodiments, the evaluation parameters may include, but are not limited to, numbers, percentages, and other manifestations. For example, if the evaluation parameter is represented by a number, the larger the number, the greater the probability that the area is an obstacle. For another example, if the evaluation parameter is expressed as a percentage, the larger the percentage, the greater the probability that the area is an obstacle. In other embodiments, the evaluation parameter can also be determined by identifying the type of obstacle, which is not specifically limited here.

It can be seen that the embodiment of the present invention determines voting information according to the depth information of the point cloud, takes into account the distance between the point cloud and the movable platform, and improves the accuracy of obstacle detection; the evaluation parameters are determined by combining the number of point clouds, and the evaluation parameters are used to determine Whether the grid area is the obstacle area can improve the accuracy of obstacle detection.

Please refer to FIG. 7, which is a schematic structural diagram of an obstacle detection device according to an embodiment of the present invention. Specifically, the obstacle detection device includes: a memory 701 and a processor 702.

In an embodiment, the obstacle detection device further includes a data interface 703, and the data interface 703 is used to transfer data information between the obstacle detection device and other devices.

The memory 701 may include a volatile memory (volatile memory); the memory 701 may also include a non-volatile memory (non-volatile memory); the memory 701 may also include a combination of the foregoing types of memories. The processor 702 may be a central processing unit (CPU). The processor 702 may further include a hardware chip. The aforementioned hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD) or a combination thereof. The aforementioned PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), or any combination thereof.

The memory 701 is used to store program instructions, and the processor 702 can call the program instructions stored in the memory 701 to perform the following steps:

Further, the size information includes the height of the movable platform and/or the safe span height, and the processor 702 filters the first point cloud to obtain the second point cloud according to the size information of the movable platform. When point cloud, it is specifically used for:

Filter the first point cloud to obtain a second point cloud according to the height of the movable platform and/or the safe span height.

Further, the two-dimensional plane includes a horizontal plane, and when the processor 702 projects the second point cloud onto the two-dimensional plane to obtain at least one projection image, it is specifically used for:

Project the second point cloud onto a horizontal plane to obtain at least one projection image.

Further, when the processor 702 determines the obstacle information of the surrounding environment where the movable platform is located according to the projection image, it is specifically configured to:

Dividing the projection image into multiple grid areas;

According to the point cloud in each grid area, it is determined whether each grid area is an obstacle area.

Further, when the processor 702 divides the projection image into multiple grid areas, it is specifically configured to:

Acquiring the type of the movable platform;

According to the type of the movable platform, the positions of the multiple grid regions are determined.

Further, the type of the movable platform includes an omnidirectional motion robot; when the processor 702 determines the positions of the multiple grid areas according to the type of the movable platform, it is specifically used for:

Acquiring the current position of the omnidirectional motion robot;

Using the current position of the omnidirectional motion robot as a geometric center, the positions of the multiple grid regions are determined.

Further, the type of the movable platform includes a non-omnidirectional motion robot; when the processor 702 determines the positions of the multiple grid areas according to the type of the movable platform, it is specifically used for:

Acquiring the current position of the non-omnidirectional motion robot;

Using the current position of the non-omnidirectional motion robot as a boundary point, the positions of the multiple grid regions are determined.

Obtaining the size of the movable platform;

According to the size of the movable platform, the size of each grid area of the plurality of grid areas is determined.

Further, when the processor 702 determines whether each grid area is an obstacle area according to the point cloud in each grid area, it is specifically configured to:

Acquiring the number of point clouds in each grid area;

When the number of the point clouds is greater than the preset number threshold, it is determined that the grid area is an obstacle area.

Acquiring the number and depth information of the point cloud in each grid area;

According to the number and depth information of the point cloud, it is determined whether the grid area is an obstacle area.

Further, when the processor 702 determines whether the grid area is an obstacle area according to the number and depth information of the point cloud, it is specifically configured to:

Determine the voting information of each point cloud in each grid area according to the depth information of the point cloud;

Determining the evaluation parameter of each grid area according to the number of point clouds and the voting information;

The evaluation parameter is compared with a preset parameter, and it is determined that a grid area where the evaluation parameter is greater than the preset parameter is the obstacle area.

Further, the grid area includes any one or more of obstacle areas, free areas, and unknown areas.

Further, when the processor 702 obtains the first point cloud corresponding to the surrounding environment where the movable platform is located, it is specifically used to:

Obtain the first point cloud corresponding to the surrounding environment where the movable platform is located through lidar.

Further, when the processor 702 obtains the first point cloud corresponding to the surrounding environment where the movable platform is located, it is specifically configured to:

Obtain the first point cloud corresponding to the surrounding environment where the movable platform is located through a camera.

Further, when the processor 702 obtains the first point cloud corresponding to the surrounding environment where the movable platform is located through a camera, it is specifically configured to:

Converting the point cloud obtained by the camera into a world coordinate system based on a preset conversion matrix to obtain a first point cloud corresponding to the surrounding environment where the movable platform is located;

Wherein, the preset conversion matrix includes an internal parameter matrix and an external parameter matrix, and the external parameter matrix includes a rotation matrix and/or a translation vector.

The embodiment of the present invention also provides a movable platform, the movable platform includes: a fuselage; a power system configured on the fuselage for providing moving power for the movable platform; and the obstacle detection device described above. In the embodiment of the present invention, the movable platform obtains the first point cloud corresponding to the surrounding environment where the movable platform is located, and filters the first point cloud according to the size information of the movable platform to obtain the second point cloud. Point cloud, and project the second point cloud onto a two-dimensional plane to obtain at least one projection image, so as to determine the obstacle information of the surrounding environment where the movable platform is located according to the at least one projection image, which reduces the calculation Complexity improves the efficiency and accuracy of obstacle detection.

An embodiment of the present invention also provides a computer-readable storage medium, the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement the method described in the embodiment of the present invention corresponding to FIG. , The device corresponding to the embodiment of the present invention described in FIG. 7 can also be implemented, which is not repeated here.

The computer-readable storage medium may be an internal storage unit of the device described in any of the foregoing embodiments, such as a hard disk or memory of the device. The computer-readable storage medium may also be an external storage device of the device, for example, a plug-in hard disk equipped on the device, a smart memory card (Smart Media Card, SMC), or a Secure Digital (SD) card. , Flash Card, etc. Further, the computer-readable storage medium may also include both an internal storage unit of the device and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by the terminal. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

The above-disclosed are only some embodiments of the present invention, which of course cannot be used to limit the scope of rights of the present invention. Therefore, equivalent changes made according to the claims of the present invention still fall within the scope of the present invention.

Claims

An obstacle detection method, characterized in that it is applied to a movable platform, and the method includes:

Acquiring the first point cloud corresponding to the surrounding environment where the movable platform is located;

Filtering the first point cloud to obtain a second point cloud according to the size information of the movable platform;

Project the second point cloud onto a two-dimensional plane to obtain at least one projection image;

According to the at least one projection image, the obstacle information of the surrounding environment where the movable platform is located is determined.
The method according to claim 1, wherein the size information includes the height of the movable platform and/or the safe span height, and the first point is determined according to the size information of the movable platform. The cloud is filtered to get the second point cloud, including:

Filter the first point cloud to obtain a second point cloud according to the height of the movable platform and/or the safe span height.
The method according to claim 1, wherein the two-dimensional plane comprises a horizontal plane, and the projecting the second point cloud onto the two-dimensional plane to obtain at least one projection image comprises:

Project the second point cloud onto a horizontal plane to obtain at least one projection image.
The method according to claim 1, wherein the determining obstacle information of the surrounding environment where the movable platform is located according to the projected image comprises:

Dividing the projection image into multiple grid areas;

According to the point cloud in each grid area, it is determined whether each grid area is an obstacle area.
The method according to claim 4, wherein the dividing the projected image into multiple grid areas comprises:

Acquiring the type of the movable platform;

According to the type of the movable platform, the positions of the multiple grid regions are determined.
The method according to claim 5, wherein the type of the movable platform includes an omnidirectional motion robot; and the determining the positions of the multiple grid areas according to the type of the movable platform includes:

Acquiring the current position of the omnidirectional motion robot;

Using the current position of the omnidirectional motion robot as a geometric center, the positions of the multiple grid regions are determined.
The method according to claim 5, wherein the type of the movable platform includes a non-omnidirectional motion robot; and the determining the positions of the multiple grid areas according to the type of the movable platform includes :

Acquiring the current position of the non-omnidirectional motion robot;

Using the current position of the non-omnidirectional motion robot as a boundary point, the positions of the multiple grid regions are determined.
The method according to claim 4, wherein the dividing the projected image into multiple grid areas comprises:

Obtaining the size of the movable platform;

According to the size of the movable platform, the size of each grid area of the plurality of grid areas is determined.
The method according to claim 4, wherein the determining whether each grid area is an obstacle area according to the point cloud in each grid area comprises:

Acquiring the number of point clouds in each grid area;

When the number of the point clouds is greater than the preset number threshold, it is determined that the grid area is an obstacle area.
The method according to claim 4, wherein the determining whether each grid area is an obstacle area according to the point cloud in each grid area comprises:

Acquiring the number and depth information of the point cloud in each grid area;

According to the number and depth information of the point cloud, it is determined whether the grid area is an obstacle area.
The method according to claim 10, wherein the determining whether the grid area is an obstacle area according to the number and depth information of the point cloud comprises:

Determine the voting information of each point cloud in each grid area according to the depth information of the point cloud;

Determining the evaluation parameter of each grid area according to the number of point clouds and the voting information;

The evaluation parameter is compared with a preset parameter, and it is determined that a grid area where the evaluation parameter is greater than the preset parameter is the obstacle area.
The method according to claim 4, wherein the grid area includes any one or more of an obstacle area, a free area, and an unknown area.
The method according to claim 1, wherein said acquiring a first point cloud corresponding to a surrounding environment where said movable platform is located comprises:

Obtain the first point cloud corresponding to the surrounding environment where the movable platform is located through lidar.
The method according to claim 1, wherein said acquiring a first point cloud corresponding to a surrounding environment where said movable platform is located comprises:

Obtain the first point cloud corresponding to the surrounding environment where the movable platform is located through a camera.
The method according to claim 14, wherein said acquiring, by a camera, a first point cloud corresponding to a surrounding environment where said movable platform is located comprises:

Converting the point cloud obtained by the camera into a world coordinate system based on a preset conversion matrix to obtain a first point cloud corresponding to the surrounding environment where the movable platform is located;

Wherein, the preset conversion matrix includes an internal parameter matrix and an external parameter matrix, and the external parameter matrix includes a rotation matrix and/or a translation vector.
An obstacle detection device, which is characterized by comprising a memory and a processor;

The memory is used to store program instructions;

The processor is configured to call the program instructions, and when the program instructions are executed, to perform the following operations:

Acquiring the first point cloud corresponding to the surrounding environment where the movable platform is located;

Filtering the first point cloud to obtain a second point cloud according to the size information of the movable platform;

Project the second point cloud onto a two-dimensional plane to obtain at least one projection image;

According to the at least one projection image, the obstacle information of the surrounding environment where the movable platform is located is determined.
The device according to claim 16, wherein the size information includes a height of the movable platform and/or a safe spanning height, and the processor performs a check on the first step according to the size information of the movable platform. When filtering one point cloud to obtain the second point cloud, it is specifically used for:

Filter the first point cloud to obtain a second point cloud according to the height of the movable platform and/or the safe span height.
The device according to claim 16, wherein the two-dimensional plane comprises a horizontal plane, and when the processor projects the second point cloud onto the two-dimensional plane to obtain at least one projection image, it is specifically used for:

Project the second point cloud onto a horizontal plane to obtain at least one projection image.
The device according to claim 16, wherein the processor is specifically configured to: when determining the obstacle information of the surrounding environment where the movable platform is located according to the projection image:

Dividing the projection image into multiple grid areas;

According to the point cloud in each grid area, it is determined whether each grid area is an obstacle area.
The device according to claim 19, wherein when the processor divides the projection image into multiple grid regions, it is specifically configured to:

Acquiring the type of the movable platform;

According to the type of the movable platform, the positions of the multiple grid regions are determined.
The device according to claim 20, wherein the type of the movable platform includes an omnidirectional motion robot; when the processor determines the positions of the multiple grid areas according to the type of the movable platform , Specifically used for:

Acquiring the current position of the omnidirectional motion robot;

Using the current position of the omnidirectional motion robot as a geometric center, the positions of the multiple grid regions are determined.
The device according to claim 20, wherein the type of the movable platform includes a non-omnidirectional motion robot; the processor determines the positions of the multiple grid areas according to the type of the movable platform When, specifically used for:

Acquiring the current position of the non-omnidirectional motion robot;

Using the current position of the non-omnidirectional motion robot as a boundary point, the positions of the multiple grid regions are determined.
The device according to claim 19, wherein when the processor divides the projection image into multiple grid regions, it is specifically configured to:

Obtaining the size of the movable platform;

According to the size of the movable platform, the size of each grid area of the plurality of grid areas is determined.
The device according to claim 19, wherein, when the processor determines whether each grid area is an obstacle area according to the point cloud in each grid area, it is specifically configured to:

Acquiring the number of point clouds in each grid area;

When the number of the point clouds is greater than the preset number threshold, it is determined that the grid area is an obstacle area.
The device according to claim 19, wherein, when the processor determines whether each grid area is an obstacle area according to the point cloud in each grid area, it is specifically configured to:

Acquiring the number and depth information of the point cloud in each grid area;

According to the number and depth information of the point cloud, it is determined whether the grid area is an obstacle area.
The device according to claim 25, wherein when the processor determines whether the grid area is an obstacle area according to the number and depth information of the point cloud, it is specifically configured to:

Determine the voting information of each point cloud in each grid area according to the depth information of the point cloud;

Determining the evaluation parameter of each grid area according to the number of point clouds and the voting information;

The evaluation parameter is compared with a preset parameter, and it is determined that a grid area where the evaluation parameter is greater than the preset parameter is the obstacle area.
The device according to claim 19, wherein the grid area includes any one or more of an obstacle area, a free area, and an unknown area.
The device according to claim 16, wherein when the processor obtains the first point cloud corresponding to the surrounding environment where the movable platform is located, it is specifically configured to:

Obtain the first point cloud corresponding to the surrounding environment where the movable platform is located through lidar.
The device according to claim 16, wherein when the processor obtains the first point cloud corresponding to the surrounding environment where the movable platform is located, it is specifically configured to:

Obtain the first point cloud corresponding to the surrounding environment where the movable platform is located through a camera.
The device according to claim 29, wherein when the processor acquires the first point cloud corresponding to the surrounding environment where the movable platform is located through a camera, it is specifically configured to:

Converting the point cloud obtained by the camera into a world coordinate system based on a preset conversion matrix to obtain a first point cloud corresponding to the surrounding environment where the movable platform is located;

Wherein, the preset conversion matrix includes an internal parameter matrix and an external parameter matrix, and the external parameter matrix includes a rotation matrix and/or a translation vector.
A movable platform, characterized in that it comprises:

body;

The power system configured on the fuselage is used to provide mobile power for the movable platform;

And the obstacle detection device according to any one of claims 16-30.
An obstacle detection system, characterized by comprising: obstacle detection equipment and a movable platform;

The obstacle detection device is used to obtain a first point cloud corresponding to the surrounding environment where the movable platform is located; filter the first point cloud to obtain a second point according to the size information of the movable platform Cloud; project the second point cloud onto a two-dimensional plane to obtain at least one projection image; determine the obstacle information of the surrounding environment where the movable platform is located according to the at least one projection image; and combine the obstacle The information is sent to the mobile platform;

The movable platform is used for moving around obstacles according to the received obstacle information.
A computer-readable storage medium storing a computer program, wherein the computer program is executed by a processor to implement the method according to any one of claims 1 to 15.