Improved DWA algorithm for epidemic prevention robot based on road guide line

Abstract

The invention relates to an improved DWA algorithm for an epidemic prevention robot based on a road guide line, and aims at solving the problem that the epidemic prevention robot can carry out obstacle avoidance path planning while following the road guide line under the environment that outdoor road conditions are complex and various obstacles exist. The method comprises the following steps: S1, performing scene analysis to obtain road semantic information; s2, image processing is carried out to extract a road area; s3, sampling road discrete points, and fitting by adopting a B spline curve to obtain a road guide line; s4, acquiring obstacle type information by using the scene analysis semantic tag; s5, binocular vision is adopted by the epidemic prevention robot, and obstacle position information is obtained; s6, the epidemic prevention robot runs an improved DWA path planning algorithm; s7, the epidemic prevention robot arrives at the target point, and a final path is obtained; compared with an existing algorithm, the method is short in operation time and low in path cost in the complex outdoor road environment, the safe distance between the method and the obstacle is always kept, the epidemic prevention robot is effectively guided to avoid the obstacle in the complex outdoor environment, and the path planning efficiency and the operation safety of the epidemic prevention robot are greatly improved.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent robot path planning, and in particular relates to an improved DWA algorithm for an epidemic prevention robot based on road guide lines. Background Art

[0002] In order to reduce the contact and infection between people, the demand for "machine replacement" has been stimulated, and epidemic prevention robots have been put on the front line of epidemic prevention in the fields of medical treatment, distribution, inspection, and household use. The epidemic prevention robots have been transformed and upgraded on the basis of the original intelligent service robots. They can effectively reduce the contact between people and reduce the risk of virus transmission. They are often used in the fields of medical treatment, distribution, inspection, and household use during the epidemic. They have functions such as temperature measurement, identification of mask wearing, health code identification, disinfection and cleaning, and assist hospitals, schools, communities and other outdoor public places in epidemic prevention and control. The epidemic prevention robots are used to complete the needs of epidemic prevention, disinfection, and medical treatment. Path planning is an important prerequisite for the autonomous navigation of epidemic prevention robots in complex outdoor environments.

[0003] The dynamic window approach (DWA) combines sensor data for online real-time local path planning. It has good obstacle avoidance capabilities and is suitable for robot path planning in dynamic environments, but it cannot meet the optimal global path. Integrating the local path planning algorithm into the global path planning, in the actual operation process, not only takes into account obstacles in the environment, but also ensures the completeness and real-time nature of the path planning and satisfies the robot's motion constraints. It can be seen that improving the safety and feasibility of path planning for epidemic prevention robots in actual scenarios is of great significance. Summary of the invention

[0004] 1. The present invention relates to an improved DWA algorithm for an epidemic prevention robot based on road guide lines, which is used to solve the problem that the epidemic prevention robot is in an outdoor road environment with complex road conditions and a variety of obstacles, so that the epidemic prevention robot can perform path planning to avoid obstacles while following the road guide lines.

[0005] Step S1: scene analysis to obtain road semantic information;

[0006] Step S2: extracting road area through image processing;

[0007] Step S3: sampling discrete points of the road and using B-spline curve fitting to obtain road guide lines;

[0008] Step S4: Obtain obstacle type information using scene parsing semantic tags;

[0009] Step S5: The epidemic prevention robot uses binocular vision to obtain obstacle location information;

[0010] Step S6: The epidemic prevention robot runs the improved DWA path planning algorithm;

[0011] Step S7: The epidemic prevention robot reaches the target point and obtains the final path;

[0012] 2. The present invention is an improved DWA algorithm for an epidemic prevention robot based on a road guide line, which is characterized in that the road guide line is used to guide the epidemic prevention robot to plan its path while avoiding obstacles, including the following steps:

[0013] Step S1: scene analysis to obtain road semantic information;

[0014] Step S2: extracting road area through image processing;

[0015] Step S3: guiding discrete point sampling, using B-spline curve fitting to obtain road guide lines;

[0016] Step S4: Obtain obstacle type information using scene parsing semantic tags;

[0017] Step S5: The epidemic prevention robot uses binocular vision to obtain obstacle location information;

[0018] Step S6: The epidemic prevention robot runs the improved DWA path planning algorithm;

[0019] Step S7: The epidemic prevention robot reaches the target point and obtains the final path;

[0020] 3. According to claim 1, the improved DWA algorithm of the epidemic prevention robot based on the road guide line comprises the following steps S1 and S2:

[0021] The outdoor road area obtained by the present invention through the semantic label map has some unclear edges and small-range pixel label errors. This paper uses a method based on morphological filtering and connected domain processing to process the outdoor road area. By performing noise reduction, filling and other operations on the outdoor road image, the accuracy of the extracted outdoor road guide line is improved. The most basic operations of morphological filtering are corrosion and filtering. By improving and combining basic operations, advanced operations such as opening operations, closing operations, morphological gradients, top hats, and black hats can be obtained. The present invention uses the cv2.morphologyEx() function to perform morphological opening and closing operations, uses a kernel size of (50,50) to remove background noise, remove burrs on the edge of the road, and improve the smoothness of the edge of the road area; uses a kernel size of (40,40) to fill in small holes in the road area and connect the cracks on the contour line. Then use the cv2.findContours() function to search for the road area contour, and finally use the cv2.drawContours() function to draw the road contour to obtain the road contour of the outdoor road. The present invention obtains the road contours of the pebbled road, the dirt road and the cement road respectively, and uses red curves to depict and mark the road areas.

[0022] 4. According to claim 2, the improved DWA algorithm of the epidemic prevention robot based on the road guide line comprises the following steps S3:

[0023] This step processes the data of the entire road profile image, which increases the computational complexity of the epidemic prevention robot. The present invention selects four-fifths of the road profile image as the region of interest (ROI). The ROI area is evenly divided into several small areas, and the road part is searched at the upper and lower edges of the small area. The center point of the road area on the sampling line is calculated to obtain the road guidance discrete point. The blue boxed area in the acquired road area is the selected ROI area, the red horizontal line represents the sampling line, the number of sampling lines used is selected as 5, and the green dots represent the obtained road guidance discrete points.

[0024] Compared with polynomial, exponential and logarithmic curves, B-spline curves have the advantages of local controllability, convex combination and continuity. In this paper, the discrete guide points generated in the ROI area are used as the control vertices of the B-spline curve, and the B-spline curve fitting equation is defined as:

[0025]

[0026] Among them, d i Represents the control points of the B-spline curve. A B-spline curve of degree n requires corresponding n+1 control points to determine the curve. i,k(t) is the basis function used in the B-spline curve calculation. The present invention uses a cubic B-spline curve to fit the outdoor road guide line. The basis function matrix of the cubic B-spline curve can be expressed by the following equation:

[0027]

[0028] Among them, B i is the B-spline analytical expression of the i-th arc, a i-1 、a i 、a i+1 and a i+2 is the discrete guide point involved in the operation. In order to make the fitted curve pass through the starting point and the target point, the two nodes a before a0 -2 with a -1 , in a n Then add two nodes a n+1 with a n+2 , to ensure the kinematic constraints at the starting point and the target point. Use B-spline curves to fit the processed road discrete guide points.

[0029] 4. According to claim 3, the improved DWA algorithm of the epidemic prevention robot based on the road guide line includes the following steps S4 and S5:

[0030] Based on the steps S1, S2 and S3, step S4 uses scene analysis semantic tags to obtain obstacle type information, and step S5 uses binocular depth images to obtain obstacle position information. By fusing these two pieces of information, three-dimensional obstacle information with fused semantics can be obtained.

[0031] 5. According to claim 4, the improved DWA algorithm of the epidemic prevention robot based on the road guide line includes the following steps S6 and S7:

[0032] Based on the steps S1, S2, S3, S4, and S5, the epidemic prevention robot has the road contour area, road guide lines, and three-dimensional information of obstacles with fusion semantics. Step S6 runs the improved DWA algorithm to process the guide lines and three-dimensional information to plan a reasonable path and guide the epidemic prevention robot to perform disinfection and temperature measurement in a complex outdoor environment. Step S8 gives the final complete path after running the improved DWA algorithm.

[0033] The present invention includes the following beneficial effects: In the present invention, inspired by the above insights, path planning in three outdoor road environments with different complexities is considered, the advantages of path guide lines are adopted to guide the epidemic prevention robot to reasonably avoid obstacles, and on this basis, combined with the improved DWA algorithm, the problem of intelligent robot path planning is deeply studied in the mean square sense.

[0034] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the technical solution of the present invention, but do not constitute a limitation to the technical solution of the present invention.

[0036] Figure 1 is a flow chart of an improved DWA algorithm of an epidemic prevention robot based on road guide lines of the present invention;

[0037] Figure 2 This is the overall block diagram of the path planning of the improved DWA algorithm of the epidemic prevention robot based on the road guide line of the present invention;

[0038] Figure 3 The invention relates to road contour extraction and guided discrete point sampling of an improved DWA algorithm for an epidemic prevention robot based on road guide lines.

[0039] Figure 4 This is a result diagram of the improved DWA algorithm of the epidemic prevention robot based on the road guide line of the present invention using a B-spline curve to fit the road guide line;

[0040] Figure 5 It is a static and dynamic path planning result diagram of the improved DWA algorithm of the epidemic prevention robot based on the road guide line of the present invention;

[0041] Figure 6 It is a path planning result diagram of a real road section of the epidemic prevention robot improved DWA algorithm based on the road guide line of the present invention; DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0043] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims or the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0044] like Figure 1As shown, the present invention provides an RRT* algorithm for epidemic prevention robots based on improved artificial potential field guidance, and the algorithm includes the following steps:

[0045] Steps S1 and S2: The outdoor road area obtained by the semantic label map has some problems such as unclear edges and small-scale pixel label errors. This paper uses a method based on morphological filtering and connected domain processing to process the outdoor road area. By performing noise reduction and filling operations on the outdoor road image, the accuracy of the extracted outdoor road guide line is improved. The most basic operations of morphological filtering are corrosion and filtering. By improving and combining basic operations, advanced operations such as opening operations, closing operations, morphological gradients, top hats, and black hats can be obtained. The present invention uses the cv2.morphologyEx() function to perform morphological opening and closing operations, uses an opening operation with a kernel size of (50,50) to remove noise points on the background, removes burrs on the edge of the road, and improves the smoothness of the edge of the road area; uses a closing operation with a kernel size of (40,40) to fill small holes in the road area and connect the cracks on the contour line. Then use the cv2.findContours() function to search for the road area contour, and finally use the cv2.drawContours() function to draw the road contour to obtain the road contour of the outdoor road. The present invention obtains the road contours of the pebbled road, the dirt road and the cement road respectively, and uses red curves to depict and mark the road areas.

[0046] Step S3: This step processes the data of the entire road profile image, which increases the computational complexity of the epidemic prevention robot. The present invention selects four-fifths of the road profile image as the region of interest (ROI). The ROI area is evenly divided into several small areas, and the road part is searched at the upper and lower edges of the small area. The center point of the road area on the sampling line is calculated to obtain the road guidance discrete point. The blue boxed area in the acquired road area is the selected ROI area, the red horizontal line represents the sampling line, the number of sampling lines used is selected as 5, and the green dots represent the obtained road guidance discrete points.

[0047] Compared with polynomial, exponential and logarithmic curves, B-spline curves have the advantages of local controllability, convex combination and continuity. In this paper, the discrete guide points generated in the ROI area are used as the control vertices of the B-spline curve, and the B-spline curve fitting equation is defined as:

[0048]

[0049] Among them, d i Represents the control points of the B-spline curve. A B-spline curve of degree n requires corresponding n+1 control points to determine the curve. i,k(t) is the basis function used in the B-spline curve calculation. The present invention uses a cubic B-spline curve to fit the outdoor road guide line. The basis function matrix of the cubic B-spline curve can be expressed by the following equation:

[0050]

[0051] Among them, B i is the B-spline analytical expression of the i-th arc, a i-1 、a i 、a i+1 and a i+2 is the discrete guide point involved in the operation. In order to make the fitted curve pass through the starting point and the target point, the two nodes a before a0 -2 with a -1 , in a n Then add two nodes a n+1 with a n+2 , to ensure the kinematic constraints at the starting point and the target point. Use B-spline curves to fit the processed road discrete guide points.

[0052] Steps S4 and S5: Based on steps S1, S2, and S3, step S4 uses scene analysis semantic tags to obtain obstacle type information, and step S5 uses binocular depth images to obtain obstacle position information. By fusing these two pieces of information, three-dimensional obstacle information with fused semantics can be obtained.

[0053] Steps S6 and S7: Based on the steps S1, S2, S3, S4, and S5, the epidemic prevention robot has the road contour area, road guide lines, and three-dimensional information of obstacles with fusion semantics. Step S6 runs the improved DWA algorithm to process the guide lines and three-dimensional information to plan a reasonable path and guide the epidemic prevention robot to perform disinfection and temperature measurement in a complex outdoor environment. Step S8 gives the final complete path after running the improved DWA algorithm.

[0054] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above-mentioned implementation mode. Technical personnel familiar with the field can also make various equivalent deformations or substitutions without violating the spirit of the present invention. These equivalent deformations or substitutions are all included in the scope defined by the claims of the present invention.

Claims

1. The present invention is an improved DWA algorithm for epidemic prevention robots based on road guide lines, characterized in that Using road guide lines to guide the epidemic prevention robot to plan its path and avoid obstacles at the same time includes the following steps: Step S1: scene analysis to obtain road semantic information; Step S2: extracting road area through image processing; Step S3: Guide the sampling of discrete points and use B-spline curve fitting to obtain road guide lines; Step S4: Obtain obstacle type information using scene parsing semantic tags; Step S5: The epidemic prevention robot uses binocular vision to obtain obstacle location information; Step S6: The epidemic prevention robot runs the improved DWA path planning algorithm; Step S7: The epidemic prevention robot reaches the target point and obtains the final path.

2. According to claim 1, the improved DWA algorithm of the epidemic prevention robot based on the road guide line comprises the following steps S1 and S2: The outdoor road area obtained by the present invention through the semantic label map has some unclear edges and small-range pixel label errors. This paper uses a method based on morphological filtering and connected domain processing to process the outdoor road area. By performing noise reduction, filling and other operations on the outdoor road image, the accuracy of the extracted outdoor road guide line is improved. The most basic operations of morphological filtering are corrosion and filtering. By improving and combining basic operations, advanced operations such as opening operations, closing operations, morphological gradients, top hats, and black hats can be obtained. The present invention uses the cv2.morphologyEx() function to perform morphological opening and closing operations, uses a kernel size of (50,50) to remove background noise, remove burrs on the edge of the road, and improve the smoothness of the edge of the road area; uses a kernel size of (40,40) to fill in small holes in the road area and connect the cracks on the contour line. Then use the cv2.findContours() function to search for the road area contour, and finally use the cv2.drawContours() function to draw the road contour to obtain the road contour of the outdoor road. The present invention obtains the road contours of the pebbled road, the dirt road and the cement road respectively, and uses red curves to depict and mark the road areas.

3. According to claim 2, the improved DWA algorithm of the epidemic prevention robot based on the road guide line comprises the following steps S3: This step processes the data of the entire road profile image, which increases the computational complexity of the epidemic prevention robot. The present invention selects four-fifths of the road profile image as the region of interest (ROI). The ROI area is evenly divided into several small areas, and the road part is searched at the upper and lower edges of the small area. The center point of the road area on the sampling line is calculated to obtain the road guidance discrete point. The blue boxed area in the acquired road area is the selected ROI area, the red horizontal line represents the sampling line, the number of sampling lines used is selected as 5, and the green dots represent the obtained road guidance discrete points. Compared with polynomial, exponential and logarithmic curves, B-spline curves have the advantages of local controllability, convex combination and continuity. In this paper, the discrete guide points generated in the ROI area are used as the control vertices of the B-spline curve, and the B-spline curve fitting equation is defined as: in, d i Represents the control points of the B-spline curve. A B-spline curve of degree n requires corresponding n+1 control points to determine the curve. i,k (t) is the basis function used in the B-spline curve calculation. The present invention uses a cubic B-spline curve to fit the outdoor road guide line. The basis function matrix of the cubic B-spline curve can be expressed by the following equation: Among them, B i is the B-spline analytical expression of the i-th arc, a i-1 、a i 、a i+1 and a i+2 is the discrete guide point involved in the operation. In order to make the fitted curve pass through the starting point and the target point, the two nodes a before a0 -2 with a -1 , in a n Then add two nodes a n+1 with a n+2 , to ensure the kinematic constraints at the starting point and the target point. Use B-spline curves to fit the processed road discrete guide points.

4. According to claim 3, the improved DWA algorithm of the epidemic prevention robot based on the road guide line includes the following steps S4 and S5: Based on the steps S1, S2 and S3, step S4 uses scene analysis semantic tags to obtain obstacle type information, and step S5 uses binocular depth images to obtain obstacle position information. By fusing these two pieces of information, three-dimensional obstacle information with fused semantics can be obtained.

5. According to claim 4, the improved DWA algorithm of the epidemic prevention robot based on the road guide line includes the following steps S6 and S7: Based on the steps S1, S2, S3, S4, and S5, the epidemic prevention robot has the road contour area, road guide lines, and three-dimensional information of obstacles with fusion semantics. Step S6 runs the improved DWA algorithm to process the guide lines and three-dimensional information to plan a reasonable path and guide the epidemic prevention robot to perform disinfection and temperature measurement in a complex outdoor environment. Step S8 gives the final complete path after running the improved DWA algorithm.

Images