A robot traffic safety assessment method and system based on obstacle spatial distribution

Abstract

The present invention discloses a robot passage safety assessment method and device based on the spatial distribution of obstacles. The method comprises the following steps: obtaining a road boundary line by collecting an image; graying the image, constructing an intuitive fuzzy set of the image, calculating the intuitive fuzzy distance, obtaining a first threshold of the minimum intuitive fuzzy distance, and identifying and segmenting obstacles; determining effective obstacles according to the relationship between the obstacles and the road boundary line, and classifying the effective obstacles; determining a passage safety level according to the spatial distribution of static effective obstacles, the movement speed of dynamic effective obstacles, and the travel speed of the robot, thereby improving the image recognition efficiency and the reliability of the robot operation.

Description

Technical Field

[0001] The present invention belongs to the field of robot control, and in particular relates to a robot passage safety assessment method and system based on obstacle spatial distribution. Background Art

[0002] At present, cleaning robots are an important part of the future urban road cleaning system. They are a comprehensive mobile robot system with environmental perception, motion planning, motion control and cleaning functions, involving multiple research fields of mobile robot technology. They have the advantages of flexibility, high efficiency and unmanned control, and can effectively improve the cleanliness of urban roads. Although cleaning robots are suitable for cleaning wide roads and community streets, the cleaning of urban road auxiliary roads is mostly done manually by sanitation workers, which is labor-intensive, with poor working conditions and low efficiency.

[0003] In the auxiliary road environment, the environmental images captured by the sweeping robot video will be affected by the weather and lighting at the time, resulting in a low obstacle avoidance accuracy rate for the outdoor sweeping robot. In addition, during the actual operation of the sweeping robot, the operating environment is partially unknown, and the obstacle distribution in the environment is dynamically uncertain and changing. How to stably drive in an environment with dynamically uncertain obstacles has become an urgent problem to be solved. Summary of the invention

[0004] In view of the low obstacle avoidance accuracy of the existing outdoor sweeping robots, the dynamic and uncertain changes in the distribution of obstacles in the environment, and the environmental image being affected by the weather and lighting at the time, the present application adopts: obtaining images in front and behind the robot according to the image acquisition module, filtering the images and performing histogram equalization to obtain road boundary lines; graying the images, constructing the intuitive fuzzy set of the images, and calculating the intuitive fuzzy distance to obtain the first threshold of the minimum intuitive fuzzy distance; using the first threshold to identify and segment obstacles; determining effective obstacles according to the relationship between the obstacles and the road boundary lines, and classifying the effective obstacles; determining the traffic safety level according to the spatial distribution of static effective obstacles, the movement speed of dynamic effective obstacles and the travel speed of the robot, which can effectively reduce the impact of light on environmental images and improve image recognition efficiency. At the same time, by identifying the distribution of different types of obstacles in the environment, the safety of the robot's operations can be well evaluated, the labor intensity of sanitation workers can be reduced, and the degree of dependence on manual labor can be reduced.

[0005] The technical solution adopted by the present invention to solve the above technical problems is: the method comprises:

[0006] A robot passage safety assessment method based on obstacle spatial distribution includes the following steps:

[0007] S1. Obtain images in front of and behind the robot according to the image acquisition module, perform histogram equalization on the images after filtering, and obtain road boundary lines;

[0008] S2, graying the image, constructing an intuitionistic fuzzy set of the image, and calculating the intuitionistic fuzzy distance to obtain a first threshold value of the minimum intuitionistic fuzzy distance;

[0009] S3, using the first threshold to identify and segment obstacles;

[0010] S4. determining effective obstacles according to the relationship between the obstacles and the road boundary line, and classifying the effective obstacles;

[0011] S5. Determine the traffic safety level according to the spatial distribution of static effective obstacles, the movement speed of dynamic effective obstacles and the travel speed of the robot;

[0012] S6. If the safety level exceeds the second safety threshold, the alarm module sends out a sound alarm message and a flash alarm message.

[0013] Furthermore, the step of performing histogram equalization after filtering the image to obtain the road boundary line includes: filtering the auxiliary road image using a Gaussian filter function or a wavelet function.

[0014] Furthermore, the histogram equalization comprises:

[0015] S11. Calculate the ratio P of different gray levels in the auxiliary road image x (i) = P (x = i) = n i / n, mapped to the interval [0,1];

[0016] S12. Calculate the cumulative distribution function:

[0017] S13, updating pixel values ​​after histogram equalization, the updating model is:

[0018]

[0019] Among them, i, j are pixel values, cdf is the cumulative distribution function, M is the pixel value of the auxiliary road length in the image, and N is the pixel value of the width of the image in the auxiliary road.

[0020] Furthermore, constructing the intuitionistic fuzzy set of the image includes:

[0021] S21, determining the image size and the grayscale value of each pixel;

[0022] S22, constructing the image intuitionistic fuzzy set:

[0023] A={(g,μ A (g))|g=0,1...,L-1};

[0024]

[0025] Where g is the gray value, p(g) is the total number of pixels with gray value g, t is the threshold to be selected, L is the gray level boundary, κ is the compactness parameter of the mean of the elements within the class, κ≥0; α is the coefficient, 0<α≤1, C is the normalization parameter, m a (t) is the grayscale mean function within the target class area, m b (t) is the grayscale mean function within the background area, μ A (·) is the membership function.

[0026] Furthermore, the first threshold for obtaining the minimum intuitive fuzzy distance includes:

[0027] Where, IFD is the first threshold, M is the pixel value of the auxiliary road length in the image, N is the pixel value of the width of the image in the auxiliary road, μ A (a ij ) represents the pixel a in the i-th row and j-th column of the grayscale image A. ij The degree of membership.

[0028] Further, using the first threshold to identify and segment the obstacle includes, after obtaining a binary image according to the first threshold, constructing a closed polygon according to the lane boundary, and if the obstacle falls within the closed polygon, determining that the obstacle is a valid obstacle.

[0029] Furthermore, determining the traffic safety level according to the spatial distribution of static effective obstacles, the movement speed of dynamic effective obstacles and the travel speed of the robot includes:

[0030] S51, calculating the ratio of the moving speed of the dynamic effective obstacle to the moving speed of the robot and the vector angle;

[0031] S52, determining the distance between the robot and the static effective obstacle and the dynamic effective obstacle according to the infrared camera module;

[0032] S53, calculating the collision probability and the collision time according to the ratio of the distance to the speed and the vector angle;

[0033] S54. Obtain a traffic safety level according to the collision probability and the collision time.

[0034] Further, after determining the collision probability and the collision time to obtain the traffic safety level, if the collision probability is higher than the third threshold and the collision time is lower than the fourth threshold, the vehicle moves toward the direction of the static effective obstacle to avoid it.

[0035] A robot passage safety assessment system based on obstacle spatial distribution, comprising:

[0036] A boundary acquisition module is used to acquire images in front of and behind the robot according to the image acquisition module, and perform histogram equalization on the images after filtering to acquire the road boundary line;

[0037] A threshold acquisition module is used to grayscale the image, construct an intuitionistic fuzzy set of the image, and calculate the intuitionistic fuzzy distance to obtain a first threshold of the minimum intuitionistic fuzzy distance;

[0038] An obstacle recognition module, used to identify and segment obstacles using the first threshold;

[0039] An obstacle classification module, used to determine effective obstacles according to the relationship between the obstacles and the road boundary line, and classify the effective obstacles;

[0040] A safety assessment module, used to determine the passage safety level according to the spatial distribution of static effective obstacles, the movement speed of dynamic effective obstacles and the travel speed of the robot;

[0041] The alarm module is used to issue sound alarm information and flash alarm information according to the safety level.

[0042] Furthermore, the system also includes a human-computer interaction module for setting avoidance strategies and avoidance actions.

[0043] The beneficial effects of the present invention are as follows:

[0044] Through image processing, the impact of light on environmental images can be effectively reduced and image recognition efficiency can be improved. At the same time, by identifying the distribution of different types of obstacles in the environment, the safety of the robot's operation can be well evaluated, the labor intensity of sanitation workers can be reduced, and the degree of dependence on manual labor can be reduced.

[0045] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above description and other purposes, features and advantages of the present invention more obvious and easy to understand, the preferred embodiments are specifically cited and described in detail as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0047] Figure 1 Flowchart of the robot traffic safety assessment method based on the spatial distribution of obstacles DETAILED DESCRIPTION

[0048] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0049] In the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be connected, detachably connected, or integrated; it can be mechanically connected or electrically connected; it can be directly connected or indirectly connected through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0050] Example 1

[0051] A robot passage safety assessment method based on obstacle spatial distribution includes the following steps:

[0052] S1. Obtain images in front of and behind the robot according to the image acquisition module, perform histogram equalization on the images after filtering, and obtain road boundary lines;

[0053] S2, graying the image, constructing an intuitionistic fuzzy set of the image, and calculating the intuitionistic fuzzy distance to obtain a first threshold value of the minimum intuitionistic fuzzy distance;

[0054] S3, using the first threshold to identify and segment obstacles;

[0055] S4. determining effective obstacles according to the relationship between the obstacles and the road boundary line, and classifying the effective obstacles;

[0056] S5. Determine the traffic safety level according to the spatial distribution of static effective obstacles, the movement speed of dynamic effective obstacles and the travel speed of the robot;

[0057] S6. If the safety level exceeds the second safety threshold, the alarm module sends out a sound alarm message and a flash alarm message.

[0058] Furthermore, the step of performing histogram equalization after filtering the image to obtain the road boundary line includes: filtering the auxiliary road image using a Gaussian filter function or a wavelet function.

[0059] Furthermore, the histogram equalization comprises:

[0060] S11. Calculate the ratio P of different gray levels in the auxiliary road image x (i) = P (x = i) = n i / n, mapped to the interval [0,1];

[0061] S12. Calculate the cumulative distribution function:

[0062] S13, updating pixel values ​​after histogram equalization, the updating model is:

[0063]

[0064] Among them, i, j are pixel values, cdf is the cumulative distribution function, M is the pixel value of the auxiliary road length in the image, and N is the pixel value of the width of the image in the auxiliary road.

[0065] Furthermore, constructing the intuitionistic fuzzy set of the image includes:

[0066] S21, determining the image size and the grayscale value of each pixel;

[0067] S22, constructing the image intuitionistic fuzzy set:

[0068] A={(g,μ A (g))|g=0,1...,L-1};

[0069]

[0070] Where g is the gray value, p(g) is the total number of pixels with gray value g, t is the threshold to be selected, L is the gray level boundary, κ is the compactness parameter of the mean of the elements within the class, κ≥0; α is the coefficient, 0<α≤1, C is the normalization parameter, m a (t) is the grayscale mean function within the target class area, m b (t) is the grayscale mean function within the background area, μ A (·) is the membership function.

[0071] Furthermore, the first threshold for obtaining the minimum intuitive fuzzy distance includes:

[0072] Where, IFD is the first threshold, M is the pixel value of the auxiliary road length in the image, N is the pixel value of the width of the image in the auxiliary road, μA (a ij ) represents the pixel a in the i-th row and j-th column of the grayscale image A. ij The degree of membership.

[0073] Further, using the first threshold to identify and segment the obstacle includes, after obtaining a binary image according to the first threshold, constructing a closed polygon according to the lane boundary, and if the obstacle falls within the closed polygon, determining that the obstacle is a valid obstacle.

[0074] Furthermore, determining the traffic safety level according to the spatial distribution of static effective obstacles, the movement speed of dynamic effective obstacles and the travel speed of the robot includes:

[0075] S51, calculating the ratio of the moving speed of the dynamic effective obstacle to the moving speed of the robot and the vector angle;

[0076] S52, determining the distance between the robot and the static effective obstacle and the dynamic effective obstacle according to the infrared camera module;

[0077] S53, calculating the collision probability and the collision time according to the ratio of the distance to the speed and the vector angle;

[0078] S54. Obtain a traffic safety level according to the collision probability and the collision time.

[0079] Further, after determining the collision probability and the collision time to obtain the traffic safety level, if the collision probability is higher than the third threshold and the collision time is lower than the fourth threshold, the vehicle moves toward the direction of the static effective obstacle to avoid it.

[0080] Example 2

[0081] A robot passage safety assessment system based on obstacle spatial distribution, comprising:

[0082] A boundary acquisition module is used to acquire images in front of and behind the robot according to the image acquisition module, and perform histogram equalization on the images after filtering to acquire the road boundary line;

[0083] A threshold acquisition module is used to grayscale the image, construct an intuitionistic fuzzy set of the image, and calculate the intuitionistic fuzzy distance to obtain a first threshold of the minimum intuitionistic fuzzy distance;

[0084] An obstacle recognition module, used to identify and segment obstacles using the first threshold;

[0085] An obstacle classification module, used to determine effective obstacles according to the relationship between the obstacles and the road boundary line, and classify the effective obstacles;

[0086] A safety assessment module, used to determine the passage safety level according to the spatial distribution of static effective obstacles, the movement speed of dynamic effective obstacles and the travel speed of the robot;

[0087] The alarm module is used to issue sound alarm information and flash alarm information according to the safety level.

[0088] Furthermore, the system also includes a human-computer interaction module for setting avoidance strategies and avoidance actions.

[0089] The advantages of the present invention are:

[0090] Through image processing, the impact of light on environmental images can be effectively reduced and image recognition efficiency can be improved. At the same time, by identifying the distribution of different types of obstacles in the environment, the safety of the robot's operation can be well evaluated, the labor intensity of sanitation workers can be reduced, and the degree of dependence on manual labor can be reduced.

[0091] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A robot traffic safety assessment method based on the spatial distribution of obstacles, characterized in that: The following steps are involved: S1. Obtain images in front of and behind the robot according to the image acquisition module, perform histogram equalization on the images after filtering, and obtain road boundary lines; S2, graying the image, constructing an intuitionistic fuzzy set of the image, and calculating the intuitionistic fuzzy distance to obtain a first threshold value of the minimum intuitionistic fuzzy distance; S3, using the first threshold to identify and segment obstacles; S4. determining effective obstacles according to the relationship between the obstacles and the road boundary line, and classifying the effective obstacles; S5. Determine the traffic safety level according to the spatial distribution of static effective obstacles, the movement speed of dynamic effective obstacles and the travel speed of the robot; S6. If the safety level exceeds the second safety threshold, the alarm module sends out a sound alarm message and a flash alarm message.

2. The robot passage safety assessment method based on obstacle spatial distribution according to claim 1 is characterized in that: The step of performing histogram equalization after filtering the image to obtain the road boundary line includes: filtering the auxiliary road image using a Gaussian filter function or a wavelet function.

3. The robot passage safety assessment method based on obstacle spatial distribution according to claim 1 is characterized in that: The histogram equalization comprises: S11. Calculate the ratio P of different gray levels in the auxiliary road image x (i) = P (x = i) = n i / n, mapped to the interval [0,1]; S12. Calculate the cumulative distribution function: S13, updating pixel values ​​after histogram equalization, the updating model is: Among them, i, j are pixel values, cdf is the cumulative distribution function, M is the pixel value of the auxiliary road length in the image, and N is the pixel value of the width of the image in the auxiliary road.

4. The robot passage safety assessment method based on obstacle spatial distribution according to claim 1 is characterized in that: The intuitionistic fuzzy set for constructing the image includes: S21, determining the image size and the grayscale value of each pixel; S22, constructing the image intuitionistic fuzzy set: A{(g,μ A (g))|g=0.1...,L-1}; Where g is the gray value, p(g) is the total number of pixels with gray value g, t is the threshold to be selected, L is the gray level boundary, κ is the compactness parameter of the mean of the elements within the class, κ≥0; α is the coefficient, 0<α≤1, C is the normalization parameter, m a (t) is the grayscale mean function within the target class area, m b (t) is the grayscale mean function within the background area, μ A (·) is the membership function.

5. The robot passage safety assessment method based on obstacle spatial distribution according to claim 1 is characterized in that: The first threshold value for obtaining the minimum intuitive fuzzy distance includes: Where, IFD is the first threshold, M is the pixel value of the auxiliary road length in the image, N is the pixel value of the width of the image in the auxiliary road, μ A (a ij ) represents the pixel a in the i-th row and j-th column of the grayscale image A. ij The degree of membership.

6. The robot passage safety assessment method based on obstacle spatial distribution according to claim 1 is characterized in that: Using the first threshold to identify and segment obstacles The method comprises: after obtaining a binary image according to a first threshold value, constructing a closed polygon according to the lane boundary, and if the obstacle falls within the closed polygon, determining that the obstacle is a valid obstacle.

7. The robot passage safety assessment method based on obstacle spatial distribution according to claim 1 is characterized in that: Determining the traffic safety level according to the spatial distribution of static effective obstacles, the movement speed of dynamic effective obstacles and the travel speed of the robot includes: S51, calculating the ratio of the moving speed of the dynamic effective obstacle to the moving speed of the robot and the vector angle; S52, determining the distance between the robot and the static effective obstacle and the dynamic effective obstacle according to the infrared camera module; S53, calculating the collision probability and the collision time according to the ratio of the distance to the speed and the vector angle; S54. Obtain a traffic safety level according to the collision probability and the collision time.

8. The robot passage safety assessment method based on obstacle spatial distribution according to claim 7 is characterized in that: After determining the collision probability and the collision time to obtain the traffic safety level, if the collision probability is higher than the third threshold and the collision time is lower than the fourth threshold, the vehicle moves toward the direction of the static effective obstacle to avoid it.

9. A robot traffic safety assessment system based on obstacle spatial distribution, characterized in that: A boundary acquisition module is used to acquire images in front of and behind the robot according to the image acquisition module, and perform histogram equalization on the images after filtering to acquire the road boundary line; A threshold acquisition module is used to grayscale the image, construct an intuitionistic fuzzy set of the image, and calculate the intuitionistic fuzzy distance to obtain a first threshold of the minimum intuitionistic fuzzy distance; An obstacle recognition module, used to identify and segment obstacles using the first threshold; An obstacle classification module, used to determine effective obstacles according to the relationship between the obstacles and the road boundary line, and classify the effective obstacles; A safety assessment module, used to determine the passage safety level according to the spatial distribution of static effective obstacles, the movement speed of dynamic effective obstacles and the travel speed of the robot; The alarm module is used to issue sound alarm information and flash alarm information according to the safety level.

10. The robot passage safety assessment system based on obstacle spatial distribution according to claim 9, characterized in that: The system also includes a human-computer interaction module for setting avoidance strategies and avoidance actions.

Images