WO2022111723A1

WO2022111723A1 - Road edge detection method and robot

Info

Publication number: WO2022111723A1
Application number: PCT/CN2021/134282
Authority: WO
Inventors: 黄寅; 张涛; 吴翔; 郭璁
Original assignee: 深圳市普渡科技有限公司
Priority date: 2020-11-30
Filing date: 2021-11-30
Publication date: 2022-06-02
Also published as: CN112486172A

Abstract

A road edge detection method and a robot. The road edge detection method comprises: acquiring depth data, a robot posture, and a topological map (101); establishing a static obstacle map according to the depth data and the robot posture (102); calculating a grayscale value of the static obstacle map (103); and calculating a road edge according to the robot posture, the topological map and the grayscale value.

Description

Road edge detection method and robot

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims the priority of the Chinese patent application with the application number 202011380356.4 and the application title "Road Edge Detection Method and Robot" filed with the China Patent Office on November 30, 2020, the entire contents of which are incorporated into this application by reference.

technical field

The present application relates to the field of Internet technologies, and in particular, to a road edge detection method and a robot.

Background technique

Mobile robots move in specific scenarios to autonomously perform tasks such as distribution, guidance, inspection, and disinfection. The above scenarios include restaurants, hotels, office buildings, hospitals, etc. In these scenarios, the robot needs to build a map and perform path planning. There will be many obstacles near the path of the robot, and the robot needs to avoid these obstacles during the movement. In the prior art, the robot detects the edge of the walking road through lidar, but there is a large error in this method, and the robot is prone to collide with obstacles.

SUMMARY OF THE INVENTION

According to various embodiments of the present application, a road edge detection method and a robot are provided. A road edge detection method, the method comprising:

Obtain depth data, robot pose, and topology map;

establishing a static obstacle map according to the depth data and the robot pose;

calculating the gray value of the static obstacle map;

A road edge is calculated according to the robot pose, the topological map and the gray value.

A robot, comprising a memory and a processor, wherein computer-readable instructions that can be run on the processor are stored in the memory, and when the processor executes the computer-readable instructions, the above-mentioned road edge detection method is implemented. step.

A computer storage medium, the computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, implements the steps of the above-mentioned road edge detection method.

The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below. Other features and advantages of the present application will be apparent from the description, drawings, and claims.

Description of drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only These are some embodiments of the present application. For those of ordinary skill in the art, the drawings of other embodiments can also be obtained according to these drawings without creative efforts.

FIG. 1 shows a schematic flowchart of the road edge detection method involved in the present application;

FIG. 2 shows a schematic flowchart of an embodiment of the road edge detection method involved in the present application;

FIG. 3 shows a schematic flowchart of an embodiment of the road edge detection method involved in the present application;

Detailed ways

In order to facilitate understanding of the present application, the present application will be described more fully below with reference to the related drawings. The preferred embodiments of the present application are shown in the accompanying drawings. However, the application may be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that a thorough and complete understanding of the disclosure of this application is provided.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field of the invention. The terms used herein in the description of the invention are for the purpose of describing particular embodiments only and are not intended to limit the application. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.

Hereinafter, preferred embodiments of the present application will be described in detail with reference to the accompanying drawings. In the following description, the same reference numerals are assigned to the same components, and overlapping descriptions are omitted. In addition, the drawings are only schematic diagrams, and the ratios of the dimensions of the members, the shapes of the members, and the like may be different from the actual ones.

As shown in FIG. 1 , the embodiment of the present application relates to a road edge detection method, including:

101. Obtain depth data, robot pose, and topology map;

102. Establish a static obstacle map according to the depth data and the robot pose;

103. Calculate the grayscale value of the static obstacle map;

104. Calculate a road edge according to the robot pose, the topological map, and the gray value.

In this case, the depth data, robot pose, topological map and static obstacle map are fused to more accurately calculate the road edge and reduce the possibility of robot collision.

In this embodiment, the depth data can be acquired by scanning the surrounding environment of the robot through the lidar set on the robot. The robot pose includes the position information and orientation information of the robot. The pose of the robot can be obtained through lidar, IMU, or odometer, etc. The topological map is the artificially set moving path of the robot.

As shown in FIG. 2, in this embodiment, after step 102, it includes:

1021. Set a plurality of measurement grids on the static obstacle map, the resolution of the measurement grids is the same as the resolution of the static obstacle map, and compare the plurality of measurement grids with the static obstacles. The map is aligned;

1022, convert the depth point cloud of the depth data from the robot coordinate system to the world coordinate system by the robot pose, and project it to the ground;

1023. Mark the measurement grid according to whether there is the depth point cloud in the measurement grid;

1024. When there is the depth point cloud in the measurement grid, the grid value of the static obstacle map in the corresponding area of the measurement grid is increased by the first feature value, and the measurement grid does not have the depth point cloud. When the depth point cloud is obtained, the grid value of the static obstacle map in the area corresponding to the measurement grid is reduced by a second feature value.

Therefore, by fusing depth point clouds and measuring grids, it is possible to quantify and distinguish grids with or without depth point clouds in the static obstacle map more significantly.

In this embodiment, step 103 specifically includes:

After the grid values of the static obstacle map of several consecutive frames are fused to increase the first eigenvalue or decrease the second eigenvalue, the gray value is calculated.

In this case, the gray value fuses the changes of the grid values of consecutive frames, so that the identification of obstacles is more accurate.

In this embodiment, the grayscale value is the grayscale value of each pixel.

In this embodiment, step 1023 specifically includes:

The measurement grid with the depth point cloud is marked as 1, and the measurement grid without the depth point cloud is marked as 0.

As shown in FIG. 3, in this embodiment, step 104 specifically includes:

1041. Find the topological path of the road where the robot is currently located in the topological map according to the robot pose;

1042. Sampling at specific spatial intervals along the topological path;

1043. Taking the sampling position as a starting point, query the gray value along the normal direction of the topological path;

1044. When the gray value is greater than the threshold, record the corresponding coordinate position;

1045. Fit several of the coordinate positions into a straight line.

In this case, when the gray value is greater than the threshold, the corresponding pixel is the peak pixel, which can be identified as the location of the obstacle. According to the location of the obstacle and the topological path, the road edge can be fitted, thereby improving the road edge. detection accuracy.

In some examples, the spatial intervals may be equally spaced.

In this embodiment, sampling is performed at specific spatial intervals, which can reduce the amount of calculation and improve the calculation efficiency.

In some examples, in one sampling, the gray value of the sampling point is [0, 0, 10, 15, 20, 200, 215, 170, 120, 180, 100, 50], and the threshold may be 190, then the coordinate positions corresponding to the two gray values of 200, 215 are recorded.

In this embodiment, step 1045 specifically includes:

The line is fitted using a random sampling consensus algorithm.

In this embodiment, the fitting of the straight line using a random sampling consensus algorithm specifically includes:

calculating the confidence of the straight line;

The straight line with the highest score on both sides of the topological map is selected as the road edge.

In this embodiment, several straight lines are fitted on both sides of the topological map. The confidence of the straight line is calculated, and the straight line that is the edge of the road is screened.

In this embodiment, the calculating the confidence level of the straight line specifically includes:

The calculation method of the confidence level is:

Wherein, n is the number of pixels covered by the straight line in the static obstacle map, and V is the pixel value of the pixel.

Therefore, the anti-interference ability of the fitting calculation can be improved, and the accuracy of the calculation can be improved.

In this embodiment, the depth data is acquired by a depth camera. The depth data includes a depth map.

In this embodiment, the topology map includes several topology paths.

In some examples, the topological map can be drawn manually. The topology map includes path information that the robot can travel. Topological paths can be straight lines.

Embodiments of the present application also relate to a robot, including a memory and a processor. Computer-readable instructions that can be run on the processor are stored in the memory. When the processor executes the computer-readable instructions, the steps of the above-mentioned road edge detection method are implemented. A depth camera can be included. The depth camera acquires the depth data. The robot may further include at least one of a lidar, an IMU, and an odometer for acquiring the robot pose.

The embodiments of the present application also relate to a computer storage medium, where the computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, implements the steps of the above-mentioned road edge detection method.

The above-mentioned embodiments do not constitute a limitation on the protection scope of the technical solution. Any modifications, equivalent replacements and improvements made within the spirit and principles of the above-mentioned embodiments shall be included within the protection scope of this technical solution.

The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity, all possible combinations of the technical features in the above-described embodiments are not described. However, as long as there is no contradiction between the combinations of these technical features, All should be regarded as the scope described in this specification.

The above-mentioned embodiments only represent several embodiments of the present application, and the descriptions thereof are relatively specific and detailed, but should not be construed as a limitation on the scope of the patent application. It should be noted that, for those skilled in the art, without departing from the concept of the present application, several modifications and improvements can be made, which all belong to the protection scope of the present application. Therefore, the scope of protection of the patent of the present application shall be subject to the appended claims.

Claims

A road edge detection method, the method comprising:

Obtain depth data, robot pose, and topology map;

establishing a static obstacle map according to the depth data and the robot pose;

calculating the gray value of the static obstacle map;

A road edge is calculated according to the robot pose, the topological map and the gray value.
The road edge detection method according to claim 1, wherein after the step of establishing a static obstacle map according to the depth data and the robot pose, the method comprises:

A plurality of measurement grids are set on the static obstacle map, the resolution of the measurement grid is the same as that of the static obstacle map, and the plurality of measurement grids are compared with the static obstacle map. position alignment;

Transform the depth point cloud of the depth data from the robot coordinate system to the world coordinate system by the robot pose, and project it to the ground;

marking the measurement grid according to whether the depth point cloud exists in the measurement grid;

When the measurement grid has the depth point cloud, the grid value of the static obstacle map in the corresponding area of the measurement grid increases the first feature value, and the measurement grid does not have the depth point. When it is cloudy, the grid value of the static obstacle map in the area corresponding to the measurement grid is reduced by a second feature value.
The road edge detection method according to claim 2, wherein the calculating the gray value of the static obstacle map specifically includes:

After the grid values of the static obstacle map of several consecutive frames are fused to increase the first eigenvalue or decrease the second eigenvalue, the gray value is calculated.
The road edge detection method according to claim 2, wherein the marking the measurement grid according to whether there is the depth point cloud in the measurement grid specifically includes:

The measurement grid with the depth point cloud is marked as 1, and the measurement grid without the depth point cloud is marked as 0.
The road edge detection method according to claim 1, wherein calculating the road edge according to the robot pose, the topological map and the grayscale value specifically includes:

Find the topological path of the road where the robot is currently located in the topological map according to the pose of the robot;

sampling at specific spatial intervals along the topological path;

Taking the sampling position as a starting point, query the gray value along the normal direction of the topological path;

When the gray value is greater than the threshold, record the corresponding coordinate position;

Fit several of the coordinate positions to a straight line.
The road edge detection method according to claim 5, wherein the fitting a plurality of the coordinate positions into a straight line specifically includes:

The line is fitted using a random sampling consensus algorithm.
The road edge detection method according to claim 6, wherein the fitting of the straight line using a random sampling consensus algorithm specifically includes:

calculating the confidence of the straight line;

The straight line with the highest score on both sides of the topological map is selected as the road edge.
The road edge detection method according to claim 7, wherein the calculating the confidence level of the straight line specifically includes:

The calculation method of the confidence level is:

Wherein, n is the number of pixels covered by the straight line in the static obstacle map, and V is the pixel value of the pixel.
The road edge detection method according to claim 1, wherein the depth data is acquired by a depth camera.
The road edge detection method according to claim 1, wherein the acquisition process of the depth data comprises:

The surrounding environment of the robot is scanned by the lidar on the robot to obtain the depth data.
The road edge detection method according to claim 1, wherein the robot pose includes position information and orientation information of the robot.
The road edge detection method according to claim 1, wherein the acquisition process of the robot pose comprises:

The robot pose is obtained by using lidar, IMU or odometer.
The road edge detection method according to claim 1, wherein the topological map is specifically a robot movement path set manually.
The road edge detection method according to claim 1, wherein the topology map includes path information that the robot can pass.
The road edge detection method according to claim 5, wherein the topological map includes several topological paths.
The road edge detection method according to claim 5, wherein the process of sampling at specific spatial intervals along the topological path comprises:

Samples are taken at equal intervals along the topological path.
The road edge detection method of claim 5, further comprising:

When the gray value is greater than the threshold, the pixel corresponding to the gray value is determined as the location of the obstacle.
The road edge detection method of claim 17, further comprising:

The road edge is fitted according to the location of the obstacle and combined with the topological path.
A robot, comprising a memory and a processor, wherein computer-readable instructions that can be executed on the processor are stored in the memory, and when the processor executes the computer-readable instructions, any one of claims 1-18 is implemented. A step of the road edge detection method.
A computer storage medium, the computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, implements the steps of the road edge detection method according to any one of claims 1-18 .