Map optimization method and device, mobile robot and storage medium

Abstract

The application relates to the technical field of image processing, and discloses a map optimization method and device, a mobile robot and a storage medium, which are used for improving the accuracy of a map. The map optimization method comprises the following steps: performing image processing on an initial map to generate at least one initial image region; performing area screening and multidimensional space processing based on the at least one initial image region to obtain an initial concave-convex point pair set corresponding to each candidate outer contour in at least one target image region; if there is at least one target initial concave-convex point pair in each candidate outer contour, the at least one target initial concave-convex point pair is added to a candidate concave-convex point pair set of the corresponding candidate outer contour; performing concave-convex point pair correction and region screening on the candidate concave-convex point pair set corresponding to each candidate outer contour to obtain at least one target segmentation region corresponding to each candidate outer contour; and generating a target map based on the at least one target segmentation region corresponding to each candidate outer contour and the initial map.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method, apparatus, mobile robot, and storage medium for map optimization. Background Technology

[0002] Currently, robotic vacuum cleaners account for a major portion of the sales of personal and household service robots, indicating that the development of robotic vacuum cleaners is progressing rapidly. As an indispensable sensor for robotic vacuum cleaners, LiDAR is also being used more and more widely.

[0003] When existing robotic vacuum cleaners use LiDAR to build maps indoors, the accuracy of the maps is low due to factors such as door gaps, wall gaps, the large field of view of LiDAR, and the insufficient precision of LiDAR. As a result, some accessible areas on the map are inaccessible to the robotic vacuum cleaner. Summary of the Invention

[0004] This invention provides a map optimization method, apparatus, mobile robot, and storage medium to improve map accuracy.

[0005] The first aspect of the present invention provides a map optimization method, comprising: acquiring an initial map to be processed, and performing image processing on the initial map to generate at least one corresponding initial image region, the at least one initial image region including multiple contour points; performing area filtering and multi-dimensional spatial processing based on the at least one initial image region and the corresponding multiple contour points to obtain at least one target image region and a set of initial concave-convex point pairs corresponding to each candidate outer contour in the at least one target image region, wherein the at least one target image region contains a region that restricts the movement of a mobile robot; if there is at least one target initial concave-convex point pair that conforms to a preset overlap filtering rule in the set of initial concave-convex point pairs corresponding to each candidate outer contour, then adding the at least one target initial concave-convex point pair to the set of candidate concave-convex point pairs of the corresponding candidate outer contour; performing concave-convex point pair correction and region filtering on the set of candidate concave-convex point pairs corresponding to each candidate outer contour to obtain at least one target segmentation region corresponding to each candidate outer contour, the at least one target segmentation region allowing the mobile robot to move; and generating a target map based on the at least one target segmentation region corresponding to each candidate outer contour and the initial map.

[0006] In one feasible implementation, the step of acquiring the initial map to be processed and performing image processing on the initial map to generate at least one corresponding initial image region includes: acquiring the initial map to be processed, the initial map including at least one reachable region, at least one obstacle region and at least one unreachable region; performing binarization processing on the initial map to obtain a binarized map; and performing an opening operation on the binarized map to generate at least one corresponding initial image region.

[0007] In one feasible implementation, the step of performing area filtering and multi-dimensional spatial processing based on the at least one initial image region and the corresponding multiple contour points to obtain at least one target image region and an initial set of concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region includes: calculating the area enclosed by each initial image region in the at least one initial image region; determining the initial image region with an area greater than or equal to a preset area as the target image region to obtain at least one target image region; and performing multi-dimensional spatial processing based on the multiple contour points of the at least one target image region to obtain an initial set of concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region.

[0008] In one feasible implementation, the step of performing multi-dimensional spatial processing based on multiple contour points of the at least one target image region to obtain an initial set of concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region includes: establishing a multi-dimensional tree structure based on multiple contour points of the at least one target image region to obtain multiple pairs of concave and convex points to be screened corresponding to each candidate outer contour in the at least one target image region, wherein the first and second pairs of concave and convex points to be screened in the pairs of concave and convex points to be screened are on different sides, and the distance between the first and second pairs of concave and convex points to be screened is the shortest distance; determining the pairs of concave and convex points to be screened that meet the preset concave and convex point pair screening rules in each candidate outer contour as initial pairs of concave and convex points to obtain multiple initial pairs of concave and convex points for each candidate outer contour; and combining the multiple initial pairs of concave and convex points for each candidate outer contour to obtain an initial set of concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region.

[0009] In one feasible implementation, the step of correcting and filtering the candidate concave / convex point pairs for each candidate outer contour to obtain at least one target segmentation region for each candidate outer contour includes: if at least two target initial concave / convex point pairs are in a positional intersection state or meet a preset deletion rule in the candidate concave / convex point pair set of each candidate outer contour, then the at least two target initial concave / convex point pairs are updated to generate at least one target concave / convex point pair for each candidate outer contour; if the number of at least one target concave / convex point pair for each candidate outer contour is one, then the candidate outer contour corresponding to the one target concave / convex point pair is segmented to obtain two candidate segmentation regions for each candidate outer contour; and then the two target concave / convex point pairs for each candidate outer contour are segmented. Candidate segmentation regions that conform to preset region rules are determined as target segmentation regions, resulting in at least one target segmentation region corresponding to each candidate outer contour. If the number of at least two target concave-convex point pairs for each candidate outer contour is at least two, the at least two target concave-convex point pairs are classified to obtain the level of each target concave-convex point pair. Based on the level of each target concave-convex point pair, the corresponding candidate outer contour is segmented to obtain at least three candidate segmentation regions corresponding to each candidate outer contour. Candidate segmentation regions that conform to the preset region rules in the at least three candidate segmentation regions of each candidate outer contour are determined as target segmentation regions, resulting in at least one target segmentation region corresponding to each candidate outer contour.

[0010] In one feasible implementation, after performing area filtering and multi-dimensional spatial processing based on the at least one initial image region and the corresponding multiple contour points to obtain at least one target image region and a set of initial concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region, the method further includes: if there is no initial concave and convex point pair that meets the preset overlap filtering rule in the set of initial concave and convex point pairs corresponding to each candidate outer contour, then determining whether the ratio of the number of contour pixels of each candidate outer contour to the number of corresponding edge obstacle pixels is greater than or equal to the preset proportion; determining the candidate outer contour with a ratio greater than or equal to the preset proportion as the target outer contour, thereby obtaining at least one target outer contour; performing image filling and dilation processing on the at least one target outer contour, and performing an AND operation with the initial map to generate a target map.

[0011] In one feasible implementation, generating a target map based on at least one target segmentation region corresponding to each candidate outer contour and the initial map includes: filling the at least one target segmentation region corresponding to each candidate outer contour with an image to obtain at least one filled segmentation region corresponding to each candidate outer contour; dilating the at least one filled segmentation region corresponding to each candidate outer contour to obtain at least one dilated segmentation region corresponding to each candidate outer contour; and performing a bitwise AND operation between the at least one dilated segmentation region corresponding to each candidate outer contour and the initial map to generate the target map.

[0012] A second aspect of the present invention provides a map optimization apparatus, comprising: an acquisition and generation module, configured to acquire an initial map to be processed and perform image processing on the initial map to generate at least one corresponding initial image region, the at least one initial image region including multiple contour points; a filtering and processing module, configured to perform area filtering and multi-dimensional spatial processing based on the at least one initial image region and the corresponding multiple contour points to obtain at least one target image region and a set of initial concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region, wherein the at least one target image region contains a region that restricts the movement of a mobile robot; a point pair addition module, configured to add the at least one target initial concave and convex point pair to the candidate concave and convex point pair set of the corresponding candidate outer contour if there is at least one target initial concave and convex point pair that conforms to a preset overlap filtering rule in the initial concave and convex point pair set corresponding to each candidate outer contour; a correction and filtering module, configured to perform concave and convex point pair correction and region filtering on the candidate concave and convex point pair set corresponding to each candidate outer contour to obtain at least one target segmentation region corresponding to each candidate outer contour, the at least one target segmentation region allowing the mobile robot to move; and a map generation module, configured to generate a target map based on the at least one target segmentation region corresponding to each candidate outer contour and the initial map.

[0013] In one feasible implementation, the acquisition and generation module is specifically used to: acquire an initial map to be processed, the initial map including at least one reachable area, at least one obstacle area and at least one inaccessible area; perform binarization processing on the initial map to obtain a binarized map; and perform an opening operation on the binarized map to generate at least one corresponding initial image area.

[0014] In one feasible implementation, the filtering processing module includes: a calculation unit for calculating the area enclosed by each initial image region in the at least one initial image region; a determination unit for determining the initial image regions with an area greater than or equal to a preset area as target image regions, thereby obtaining at least one target image region; and a processing unit for performing multi-dimensional spatial processing based on multiple contour points of the at least one target image region, thereby obtaining an initial set of concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region.

[0015] In one feasible implementation, the processing unit is specifically used to: establish a multi-dimensional tree structure based on multiple contour points of the at least one target image region, to obtain multiple pairs of concave and convex points to be screened corresponding to each candidate outer contour in the at least one target image region, wherein the first and second concave and convex points to be screened in the pairs of concave and convex points to be screened are on different sides, and the distance between the first and second concave and convex points to be screened is the shortest distance; determine the pairs of concave and convex points to be screened that meet the preset concave and convex point pair screening rules in each candidate outer contour as initial pairs of concave and convex points, to obtain multiple initial pairs of concave and convex points for each candidate outer contour; and combine the multiple initial pairs of concave and convex points for each candidate outer contour to obtain a set of initial pairs of concave and convex points corresponding to each candidate outer contour in the at least one target image region.

[0016] In one feasible implementation, the correction and filtering module is specifically used for: if at least two target initial concave-convex point pairs exist in the candidate concave-convex point pair set of each candidate outer contour in a positional intersection state or meet a preset deletion rule, then the at least two target initial concave-convex point pairs are updated to generate at least one target concave-convex point pair corresponding to each candidate outer contour; if the number of at least one target concave-convex point pair for each candidate outer contour is one, then the candidate outer contour corresponding to the one target concave-convex point pair is segmented to obtain two candidate segmentation regions for each candidate outer contour; and the candidate segmentation regions that meet the preset region rules in the two candidate segmentation regions of each candidate outer contour are determined. For each candidate outer contour, at least one target segmentation region is obtained. If the number of at least two target concave-convex point pairs for each candidate outer contour is at least two, the at least two target concave-convex point pairs are classified to obtain the level of each target concave-convex point pair. Based on the level of each target concave-convex point pair, the corresponding candidate outer contour is segmented to obtain at least three candidate segmentation regions for each candidate outer contour. The candidate segmentation regions that conform to the preset region rules among the at least three candidate segmentation regions of each candidate outer contour are determined as target segmentation regions, thus obtaining at least one target segmentation region for each candidate outer contour.

[0017] In one feasible implementation, the map optimization device further includes: a judgment module, configured to determine whether the ratio of the number of contour pixels of each candidate outer contour to the number of corresponding edge obstacle pixels is greater than or equal to the preset proportion if there is no initial concave-convex point pair that meets the preset overlap filtering rule in the initial concave-convex point pair set corresponding to each candidate outer contour; a determination module, configured to determine the candidate outer contours with a ratio greater than or equal to the preset proportion as target outer contours, thereby obtaining at least one target outer contour; and an AND operation module, configured to perform image filling and dilation processing on the at least one target outer contour, and perform an AND operation with the initial map to generate a target map.

[0018] In one feasible implementation, the map generation module is specifically used to: fill at least one target segmentation region corresponding to each candidate outer contour with an image to obtain at least one filled segmentation region corresponding to each candidate outer contour; dilate at least one filled segmentation region corresponding to each candidate outer contour to obtain at least one dilated segmentation region corresponding to each candidate outer contour; and perform a bitwise AND operation between at least one dilated segmentation region corresponding to each candidate outer contour and the initial map to generate a target map.

[0019] A third aspect of the present invention provides a mobile robot, the mobile robot comprising: a memory and at least one processor, the memory storing instructions; the at least one processor calling the instructions in the memory to cause the mobile robot to execute the map optimization method described above.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed on the mobile robot, cause the mobile robot to perform the map optimization method described above.

[0021] In the technical solution provided by this invention, an initial map to be processed is obtained, and the initial map is image processed to generate at least one corresponding initial image region, which includes multiple contour points; based on the at least one initial image region and the corresponding multiple contour points, area filtering and multi-dimensional spatial processing are performed to obtain at least one target image region and a set of initial concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region, where at least one target image region contains a region that restricts the movement of the mobile robot; if there is at least one target initial concave and convex point pair that meets the preset overlap filtering rules in the set of initial concave and convex point pairs corresponding to each candidate outer contour, the at least one target initial concave and convex point pair is added to the set of candidate concave and convex point pairs of the corresponding candidate outer contour; concave and convex point pair correction and region filtering are performed on the set of candidate concave and convex point pairs corresponding to each candidate outer contour to obtain at least one target segmentation region corresponding to each candidate outer contour, where at least one target segmentation region allows the mobile robot to move; based on the at least one target segmentation region corresponding to each candidate outer contour and the initial map, a target map is generated. In this embodiment of the invention, image processing is performed on the initial map to obtain at least one initial image region. The at least one initial image region is then subjected to area filtering, multi-dimensional spatial processing, concave-convex point pair correction, and region filtering to obtain at least one target segmentation region. Based on the at least one target segmentation region and the initial map, a target map is generated, thereby deleting reachable areas on the map that the mobile robot cannot reach. This makes the map look more aesthetically pleasing, improves the accuracy of mobile robot positioning and path planning, and enhances the accuracy of the map. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of one embodiment of the map optimization method in this invention;

[0023] Figure 2 This is a schematic diagram of another embodiment of the map optimization method in this invention;

[0024] Figure 3 This is a schematic diagram of an embodiment of the initial image region in this invention;

[0025] Figure 4 This is a schematic diagram of one embodiment of the concave and convex points in the present invention;

[0026] Figure 5 This is a schematic diagram of one embodiment of the concave-convex point pair in the present invention;

[0027] Figure 6 This is a schematic diagram of an embodiment of the present invention, showing an enlarged initial image region.

[0028] Figure 7This is a schematic diagram of an embodiment of the present invention in which the positions of the concave and convex points intersect;

[0029] Figure 8 This is a schematic diagram of an embodiment of the present invention where the concave and convex point pairs conform to a preset deletion rule;

[0030] Figure 9 This is a schematic diagram of an embodiment of the updated concave-convex point pair with intersecting positions in this invention.

[0031] Figure 10 This is a schematic diagram of an embodiment of the invention after updating the bump and concave point pairs that conform to the preset deletion rules;

[0032] Figure 11 This is a schematic diagram of an embodiment of the present invention showing the candidate outer contour of a pair of concave and convex points before segmentation;

[0033] Figure 12 This is a schematic diagram of an embodiment of the present invention showing the candidate outer contour of a pair of concave and convex points after segmentation.

[0034] Figure 13 This is a schematic diagram of an embodiment of the candidate segmentation region and edge obstacles in this invention;

[0035] Figure 14 This is a schematic diagram of an embodiment of the candidate outer contour of two pairs of concave and convex points before segmentation in this invention.

[0036] Figure 15 This is a schematic diagram of an embodiment of the present invention showing the candidate outer contours of two pairs of concave and convex points after segmentation.

[0037] Figure 16 This is a schematic diagram of one embodiment of the target segmentation region in this invention;

[0038] Figure 17 This is a schematic diagram of one embodiment of the target map in this invention;

[0039] Figure 18 This is a schematic diagram of an embodiment of the present invention in which two obstacles exist between the pairs of concave and convex points;

[0040] Figure 19 This is a schematic diagram of an embodiment of the present invention showing an overlapping region between two segmented regions formed by the concave and convex points;

[0041] Figure 20 This is a schematic diagram of one embodiment of the map optimization device in this invention;

[0042] Figure 21 This is a schematic diagram of another embodiment of the map optimization device in this invention;

[0043] Figure 22 This is a schematic diagram of one embodiment of the mobile robot in this invention. Detailed Implementation

[0044] This invention provides a map optimization method, apparatus, mobile robot, and storage medium to improve map accuracy.

[0045] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” or “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0046] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the map optimization method in this invention includes:

[0047] 101. Obtain the initial map to be processed, and perform image processing on the initial map to generate at least one corresponding initial image region, wherein at least one initial image region includes multiple contour points;

[0048] It is understood that the executing entity of this invention can be a map optimization device or a mobile robot, and no specific limitation is made here. This embodiment of the invention will be described using a mobile robot as an example.

[0049] The initial map is a radar and laser map of the indoor environment. The initial map includes accessible areas, obstacle areas, and inaccessible areas. The pixel values ​​of each pixel in the accessible, obstacle, and inaccessible areas are different, and the mobile robot can only move within the accessible areas.

[0050] Image processing includes binarization and opening operations. The mobile robot binarizes the initial map, making accessible areas the foreground and obstacle and inaccessible areas the background. The mobile robot then performs morphological opening operations on the binarized initial map to eliminate the connections between accessible areas and remove spiky points in the accessible areas.

[0051] 102. Based on at least one initial image region and the corresponding multiple contour points, area filtering and multi-dimensional spatial processing are performed to obtain at least one target image region and at least one set of initial concave and convex point pairs corresponding to each candidate outer contour in the target image region. At least one target image region contains a region that restricts the movement of the mobile robot.

[0052] Mobile robots use k-dimensional trees (kd-trees) for multidimensional spatial processing. A kd-tree is a tree-like data structure that stores instance points in a k-dimensional space for fast retrieval. It is used for searching key data in multidimensional space, such as range search and nearest neighbor search.

[0053] 103. If there is at least one target initial concave-convex point pair that meets the preset overlap filtering rules in the initial concave-convex point pair set corresponding to each candidate outer contour, then add at least one target initial concave-convex point pair to the candidate concave-convex point pair set of the corresponding candidate outer contour.

[0054] The preset overlap filtering rule is used to indicate that there is no overlap between the initial concave and convex points and the two divided regions. If there is no overlap between the initial concave and convex points and the two divided regions, the mobile robot determines that there is a narrow place that restricts the movement of the mobile robot with respect to the corresponding candidate outer contour of the initial concave and convex point. If there is an overlap between the initial concave and convex points and the two divided regions, the mobile robot determines that there is no narrow place that restricts the movement of the mobile robot with respect to the corresponding candidate outer contour of the initial concave and convex point.

[0055] 104. Perform concave-convex point pair correction and region filtering on the set of candidate concave-convex point pairs corresponding to each candidate outer contour to obtain at least one target segmentation region corresponding to each candidate outer contour. At least one target segmentation region allows the mobile robot to move.

[0056] The mobile robot will update the initial concave-convex point pairs that intersect in position, and update the initial concave-convex point pairs that meet the preset deletion rules. The preset deletion rules are used to indicate that the distance between the same side of the initial concave-convex point pairs is less than a preset distance. The preset distance can be the diameter of the mobile robot, or other distances smaller than the diameter of the mobile robot.

[0057] 105. Generate a target map based on at least one target segmentation region corresponding to each candidate outer contour and an initial map.

[0058] The mobile robot performs image filling processing on the target segmented region, with the filled pixel values ​​being the same as the pixel values ​​of the reachable region. The filled target segmented region is then dilated to thin out obstacles at the edges of the reachable region, such as thinning out walls at the edges of the reachable region.

[0059] In this embodiment of the invention, image processing is performed on the initial map to obtain at least one initial image region. The at least one initial image region is then subjected to area filtering, multi-dimensional spatial processing, concave-convex point pair correction, and region filtering to obtain at least one target segmentation region. Based on the at least one target segmentation region and the initial map, a target map is generated, thereby deleting reachable areas on the map that the mobile robot cannot reach. This makes the map look more aesthetically pleasing, improves the accuracy of mobile robot positioning and path planning, and enhances the accuracy of the map.

[0060] Please see Figure 2 Another embodiment of the map optimization method in this invention includes:

[0061] 201. Obtain the initial map to be processed, and perform image processing on the initial map to generate at least one corresponding initial image region, wherein at least one initial image region includes multiple contour points;

[0062] Specifically, (1) the mobile robot acquires an initial map to be processed, the initial map including at least one reachable area, at least one obstacle area and at least one unreachable area; (2) the mobile robot performs binarization processing on the initial map to obtain a binarized map; (3) the mobile robot performs opening operation on the binarized map to generate at least one corresponding initial image area.

[0063] At least one reachable region is used to represent a pre-defined area that the mobile robot can move within. For example, the mobile robot acquires an initial map to be processed. The initial map includes at least one reachable region, at least one obstacle region, and at least one unreachable region. The at least one reachable region represents the area that the mobile robot can move within. The mobile robot performs binarization processing on the initial map to obtain a binarized map. The mobile robot performs an opening operation on the binarized map to generate at least one corresponding initial image region, such as... Figure 3 As shown.

[0064] 202. Calculate the area enclosed by each initial image region in at least one initial image region;

[0065] For example, the mobile robot calculates that the area enclosed by an initial image region is 2 square meters, or the mobile robot calculates that the area enclosed by an initial image region is 0.5 square meters.

[0066] 203. Determine the initial image region with an area greater than or equal to the preset area as the target image region, and obtain at least one target image region;

[0067] The preset area is used to represent the minimum movement area corresponding to the preset mobile robot. The preset area is the minimum movement area set for a specific mobile robot. For example, if the preset area is 1 square meter, the minimum movement area of ​​the mobile robot is 1 square meter. The mobile robot determines the initial image area with an area greater than or equal to 1 square meter as the target image area, and the mobile robot discards the initial image area with an area less than 1 square meter.

[0068] 204. Perform multi-dimensional spatial processing on multiple contour points of at least one target image region to obtain an initial set of concave and convex point pairs corresponding to each candidate outer contour in at least one target image region;

[0069] Mobile robots determine the protrusions and concave points using vector angles, such as... Figure 4 As shown, the mobile robot determines whether point a is a concave or convex point by the angle formed by vectors ab and ac. When the angle is less than or equal to a set threshold, such as 150 degrees, point a is determined to be a concave or convex point because the angle formed by any point on a straight line is 180 degrees, which does not satisfy the characteristics of a concave or convex point.

[0070] Specifically, (1) the mobile robot establishes a multidimensional tree structure based on multiple contour points of at least one target image region to obtain multiple pairs of concave and convex points to be screened for each candidate outer contour in at least one target image region. The first and second concave and convex points to be screened in the pair of concave and convex points to be screened are on different sides, and the distance between the first and second concave and convex points to be screened is the shortest distance; (2) the mobile robot determines the pairs of concave and convex points to be screened that meet the preset concave and convex point pair screening rules in each candidate outer contour as initial pairs of concave and convex points to obtain multiple initial pairs of concave and convex points for each candidate outer contour; (3) the mobile robot combines the multiple initial pairs of concave and convex points for each candidate outer contour to obtain a set of initial pairs of concave and convex points corresponding to each candidate outer contour in at least one target image region.

[0071] The preset bump pair filtering rules are used to indicate that the distance between the bumps to be filtered in a bump pair is less than a preset distance. Different sides are defined as follows: the number of pixels corresponding to the first and second bumps to be filtered is less than or equal to a preset number; or the contour length between the first and second bumps to be filtered is less than or equal to a preset contour length; or the area enclosed by the first and second bumps to be filtered is less than or equal to a preset area.

[0072] For example, a mobile robot establishes a multidimensional tree structure based on multiple contour points of at least one target image region, obtaining multiple pairs of concave and convex points to be screened for each candidate outer contour in at least one target image region. The first and second concave and convex points in each pair are on different sides, and the distance between them is the shortest distance. The mobile robot determines the pairs of concave and convex points that meet the preset screening rules for each candidate outer contour as initial pairs, thus obtaining multiple initial pairs of concave and convex points for each candidate outer contour. Figure 5 As shown, initial bump pairs ab and cd are defined. A preset bump pair filtering rule is used to indicate that the distance between the bumps to be filtered in the bump pair is less than a preset distance. The preset distance can be the diameter of the mobile robot or other distances smaller than the diameter of the mobile robot. The mobile robot combines multiple initial bump pairs for each candidate outer contour to obtain at least one set of initial bump pairs corresponding to each candidate outer contour in the target image region.

[0073] 205. If there is at least one target initial concave-convex point pair that meets the preset overlap filtering rules in the initial concave-convex point pair set corresponding to each candidate outer contour, then add at least one target initial concave-convex point pair to the candidate concave-convex point pair set of the corresponding candidate outer contour.

[0074] The preset overlap filtering rule is used to indicate that there is no overlap between the initial concave and convex points and the two divided regions. If there is no overlap between the initial concave and convex points and the two divided regions, the mobile robot determines that there is a narrow place that restricts the movement of the mobile robot with respect to the corresponding candidate outer contour of the initial concave and convex point. If there is an overlap between the initial concave and convex points and the two divided regions, the mobile robot determines that there is no narrow place that restricts the movement of the mobile robot with respect to the corresponding candidate outer contour of the initial concave and convex point.

[0075] For example, Figure 3 The initial image region was magnified to obtain Figure 6 , Figure 6 Each candidate outer contour includes at least one target initial concave-convex point pair.

[0076] 206. Perform concave-convex point pair correction and region filtering on the set of candidate concave-convex point pairs corresponding to each candidate outer contour to obtain at least one target segmentation region corresponding to each candidate outer contour. At least one target segmentation region allows the mobile robot to move.

[0077] Specifically, (1) if at least two initial target concave-convex point pairs exist in the candidate concave-convex point pair set of each candidate outer contour, and they are in a positional intersection state or meet the preset deletion rules, then the mobile robot updates at least two initial target concave-convex point pairs to generate at least one target concave-convex point pair corresponding to each candidate outer contour; (2) if the number of at least one target concave-convex point pair for each candidate outer contour is one, then the mobile robot segments the candidate outer contour corresponding to one target concave-convex point pair to obtain two candidate segmentation regions for each candidate outer contour; (3) the mobile robot determines the candidate segmentation region that meets the preset region rules in the two candidate segmentation regions of each candidate outer contour as the target segmentation region, and obtains (4) If the number of at least two target concave-convex point pairs for each candidate outer contour is at least two, the mobile robot classifies the at least two target concave-convex point pairs to obtain the level of each target concave-convex point pair in the at least two target concave-convex point pairs; (5) The mobile robot segments the corresponding candidate outer contour based on the level of each target concave-convex point pair to obtain at least three candidate segmentation regions corresponding to each candidate outer contour; (6) The mobile robot determines the candidate segmentation regions that conform to the preset region rules in the at least three candidate segmentation regions of each candidate outer contour as target segmentation regions to obtain at least one target segmentation region corresponding to each candidate outer contour.

[0078] The preset deletion rule is used to indicate that the distance between the initial concave and convex points of the target on the same side is less than the preset distance. The preset region rule is used to indicate that the ratio of the number of contour pixels of the candidate segmentation region to the number of corresponding edge obstacle pixels is greater than or equal to the preset proportion, and the area corresponding to the candidate segmentation region is greater than or equal to the preset area. The preset area is used to indicate the minimum movement area corresponding to the preset mobile robot.

[0079] For example, if there are two target initial concave-convex point pairs in the candidate concave-convex point pair set of each candidate outer contour, such as Figure 7 As shown, the initial concave-convex point pair ab and the initial concave-convex point pair cd are in a positional intersection state, or meet the preset deletion rules, such as... Figure 8 As shown, if the initial concave-convex point pair ab and the initial concave-convex point pair cd meet the preset deletion rules, the mobile robot updates the two initial concave-convex point pairs to generate a target concave-convex point pair corresponding to each candidate outer contour, as shown. Figure 9 As shown, the pairs of concave and convex points that intersect are updated to the target pair of concave and convex points ab, as follows. Figure 10As shown, bump pairs that meet the preset deletion rules are updated to target bump pairs ab or cd. The preset deletion rules indicate that the distance between the same side of the target initial bump pairs is less than a preset distance. The condition of the same side refers to the number of pixels corresponding to bump a in the target initial bump pair ab and bump c in the target initial bump pair cd being less than or equal to a preset number, or the contour length between bump a and bump c being less than or equal to a preset contour length, or the area enclosed by bump a and bump c being less than or equal to a preset area. The preset distance can be the diameter of the mobile robot or other distances smaller than the diameter of the mobile robot. If the number of at least one target bump pair for each candidate outer contour is one, the mobile robot segments the candidate outer contour according to a target bump pair, obtaining two candidate segmentation regions for each candidate outer contour, such as... Figure 11 Before and as shown Figure 12 After segmentation as shown, the mobile robot determines the candidate segmentation regions that conform to the preset region rules in the two candidate segmentation regions of each candidate outer contour as the target segmentation regions, thus obtaining at least one target segmentation region corresponding to each candidate outer contour. The preset region rules are used to indicate that the ratio of the contour pixel count to the corresponding edge obstacle pixel count in the candidate segmentation region is greater than or equal to a preset proportion, such as... Figure 13 As shown, the preset percentage can be 90%, or other values, depending on the actual situation. This is not limited here. The area corresponding to the candidate segmentation region is greater than or equal to the preset area. The preset area represents the minimum movement area corresponding to the preset mobile robot. If the number of at least one pair of target bumps / concave points for each candidate outer contour is two, the mobile robot classifies the two pairs of target bumps / concave points to obtain the hierarchy of each pair, as shown below. Figure 14 As shown, since the bumps and dents in the contour are stored in a counter-clockwise order, the target bump and dent pair ab is the parent and the target bump and dent pair cd is the child. The mobile robot segments the corresponding candidate outer contour based on the hierarchy of each target bump and dent pair, obtaining three candidate segmentation regions for each candidate outer contour, as shown below. Figure 15 As shown, the mobile robot determines the candidate segmentation regions that conform to the preset region rules in the three candidate segmentation regions of each candidate outer contour as the target segmentation regions, thus obtaining at least one target segmentation region corresponding to each candidate outer contour, as shown below. Figure 16 As shown, Figure 16 It is by Figure 6 After correction of concave and convex points and region filtering, at least one target segmentation region is obtained for each candidate outer contour.

[0080] 207. Generate a target map based on at least one target segmentation region corresponding to each candidate outer contour and an initial map.

[0081] The mobile robot performs image filling processing on the target segmented region, with the filled pixel values ​​being the same as the pixel values ​​of the reachable region. The filled target segmented region is then dilated to thin out obstacles at the edges of the reachable region, such as thinning out walls at the edges of the reachable region.

[0082] Specifically, (1) the mobile robot fills the image of at least one target segmentation region corresponding to each candidate outer contour to obtain at least one filled segmentation region corresponding to each candidate outer contour; (2) the mobile robot dilates the at least one filled segmentation region corresponding to each candidate outer contour to obtain at least one dilated segmentation region corresponding to each candidate outer contour; (3) the mobile robot performs a bitwise AND operation between the at least one dilated segmentation region corresponding to each candidate outer contour and the initial map to generate the target map.

[0083] For example, the mobile robot fills at least one target segmentation region corresponding to each candidate outer contour, obtaining at least one filled segmentation region corresponding to each candidate outer contour. The mobile robot then dilates this at least one filled segmentation region corresponding to each candidate outer contour, obtaining at least one dilated segmentation region corresponding to each candidate outer contour. Finally, the mobile robot performs a bitwise AND operation between this at least one dilated segmentation region corresponding to each candidate outer contour and the initial map to generate the target map. Figure 17 The target map shown.

[0084] In one feasible implementation, (1) the mobile robot establishes a multidimensional tree structure based on multiple contour points of at least one target image region to obtain multiple pairs of concave and convex points to be screened corresponding to each candidate outer contour in at least one target image region. The first and second concave and convex points to be screened in the pair are on different sides, and the distance between the first and second concave and convex points to be screened is the shortest distance; (2) the mobile robot determines the pairs of concave and convex points to be screened in each candidate outer contour that meet the preset concave and convex point pair screening rules or preset obstacle rules as the initial concave and convex points. Point pairs are obtained to obtain multiple initial concave and convex point pairs for each candidate outer contour. The preset concave and convex point pair filtering rules are used to indicate that the distance between the concave and convex points to be filtered in the concave and convex point pairs is less than the preset distance. The preset obstacle rules are used to indicate that there is at least one obstacle between the concave and convex point pairs to be filtered, and the distance between at least one obstacle and each concave and convex point to be filtered in the concave and convex point pairs to be filtered is less than the preset distance. (3) The mobile robot combines multiple initial concave and convex point pairs for each candidate outer contour to obtain at least one set of initial concave and convex point pairs corresponding to each candidate outer contour in the target image region.

[0085] For example, a mobile robot constructs a multidimensional tree structure based on multiple contour points of at least one target image region, obtaining multiple pairs of unselected bump points corresponding to each candidate outer contour in at least one target image region. The first and second unselected bump points in each pair are on different sides, and the distance between them is the shortest distance. The mobile robot determines the unselected bump point pairs in each candidate outer contour that meet preset bump point pair selection rules or preset obstacle rules as initial bump point pairs, obtaining multiple initial bump point pairs for each candidate outer contour. The preset bump point pair selection rules indicate that the distance between the unselected bump points in the pair is less than a preset distance. The preset distance can be the diameter of the mobile robot or other distances smaller than the diameter of the mobile robot. The preset obstacle rules indicate that there is at least one obstacle between the unselected bump point pairs, and the distance between the at least one obstacle and each unselected bump point in the pair is less than the preset distance. Figure 18 As shown, at least two obstacles are present. The mobile robot combines multiple initial bump and concave point pairs for each candidate outer contour to obtain at least one set of initial bump and concave point pairs corresponding to each candidate outer contour in the target image region.

[0086] In one feasible implementation, (1) if there is no initial concave-convex point pair that meets the preset overlap filtering rule in the initial concave-convex point pair set corresponding to each candidate outer contour, the mobile robot determines whether the ratio of the number of contour pixels of each candidate outer contour to the number of corresponding edge obstacle pixels is greater than or equal to the preset proportion; (2) the mobile robot determines the candidate outer contour with the ratio greater than or equal to the preset proportion as the target outer contour, and obtains at least one target outer contour; (3) the mobile robot performs image filling and dilation processing on at least one target outer contour, and performs AND operation with the initial map to generate the target map.

[0087] For example, if there is no initial concave-convex point pair in the set of initial concave-convex point pairs corresponding to each candidate outer contour that meets the preset overlap filtering rules, such as Figure 19 As shown, if there is an overlapping area between the two regions segmented by the initial concave and convex points (ab), the mobile robot determines whether the ratio of the number of contour pixels to the number of corresponding edge obstacle pixels of each candidate outer contour is greater than or equal to a preset percentage. The preset percentage can be 90% or other values, depending on the actual situation, and is not limited here. The mobile robot determines the candidate outer contours with a ratio greater than or equal to the preset percentage as the target outer contours, obtaining at least one target outer contour. The mobile robot performs image filling and dilation processing on at least one target outer contour and performs a bitwise AND operation with the initial map to generate the target map.

[0088] In this embodiment of the invention, image processing is performed on the initial map to obtain at least one initial image region. The at least one initial image region is then subjected to area filtering, multi-dimensional spatial processing, concave-convex point pair correction, and region filtering to obtain at least one target segmentation region. Based on the at least one target segmentation region and the initial map, a target map is generated, thereby deleting reachable areas on the map that the mobile robot cannot reach. This makes the map look more aesthetically pleasing, improves the accuracy of mobile robot positioning and path planning, and enhances the accuracy of the map.

[0089] The map optimization method in the embodiments of the present invention has been described above. The map optimization apparatus in the embodiments of the present invention will be described below. Please refer to [link / reference]. Figure 20 One embodiment of the map optimization device in this invention includes:

[0090] The generation module 2001 is used to acquire the initial map to be processed, and to perform image processing on the initial map to generate at least one corresponding initial image region, wherein the at least one initial image region includes multiple contour points.

[0091] The filtering processing module 2002 is used to perform area filtering and multi-dimensional spatial processing based on at least one initial image region and corresponding multiple contour points to obtain at least one target image region and at least one set of initial concave and convex point pairs corresponding to each candidate outer contour in the target image region, wherein at least one target image region contains a region that restricts the movement of the mobile robot.

[0092] The point pair addition module 2003 is used to add at least one target initial concave-convex point pair to the candidate concave-convex point pair set of the corresponding candidate outer contour if there is at least one target initial concave-convex point pair that meets the preset overlap filtering rules in the initial concave-convex point pair set corresponding to each candidate outer contour.

[0093] The correction and filtering module 2004 is used to correct the concave and convex point pairs and filter the region for each candidate outer contour corresponding to the set of candidate concave and convex point pairs, so as to obtain at least one target segmentation region corresponding to each candidate outer contour, and at least one target segmentation region allows the mobile robot to move.

[0094] The map generation module 2005 is used to generate a target map based on at least one target segmentation region corresponding to each candidate outer contour and an initial map.

[0095] In this embodiment of the invention, image processing is performed on the initial map to obtain at least one initial image region. The at least one initial image region is then subjected to area filtering, multi-dimensional spatial processing, concave-convex point pair correction, and region filtering to obtain at least one target segmentation region. Based on the at least one target segmentation region and the initial map, a target map is generated, thereby deleting reachable areas on the map that the mobile robot cannot reach. This makes the map look more aesthetically pleasing, improves the accuracy of mobile robot positioning and path planning, and enhances the accuracy of the map.

[0096] Please see Figure 21 Another embodiment of the map optimization device in this invention includes:

[0097] The generation module 2001 is used to acquire the initial map to be processed, and to perform image processing on the initial map to generate at least one corresponding initial image region, wherein the at least one initial image region includes multiple contour points.

[0098] The filtering processing module 2002 is used to perform area filtering and multi-dimensional spatial processing based on at least one initial image region and corresponding multiple contour points to obtain at least one target image region and at least one set of initial concave and convex point pairs corresponding to each candidate outer contour in the target image region, wherein at least one target image region contains a region that restricts the movement of the mobile robot.

[0099] The point pair addition module 2003 is used to add at least one target initial concave-convex point pair to the candidate concave-convex point pair set of the corresponding candidate outer contour if there is at least one target initial concave-convex point pair that meets the preset overlap filtering rules in the initial concave-convex point pair set corresponding to each candidate outer contour.

[0100] The correction and filtering module 2004 is used to correct the concave and convex point pairs and filter the region for each candidate outer contour corresponding to the set of candidate concave and convex point pairs, so as to obtain at least one target segmentation region corresponding to each candidate outer contour, and at least one target segmentation region allows the mobile robot to move.

[0101] The map generation module 2005 is used to generate a target map based on at least one target segmentation region corresponding to each candidate outer contour and an initial map.

[0102] Optionally, the acquisition module 2001 is specifically used for:

[0103] Obtain the initial map to be processed, which includes at least one reachable area, at least one obstacle area, and at least one unreachable area;

[0104] The initial map is binarized to obtain a binarized map;

[0105] Perform an opening operation on the binarized map to generate at least one corresponding initial image region.

[0106] Optionally, the filtering processing module 2002 includes:

[0107] The calculation unit 20021 is used to calculate the area enclosed by each initial image region in at least one initial image region;

[0108] The determining unit 20022 is used to determine an initial image region with an area greater than or equal to a preset area as a target image region, thereby obtaining at least one target image region.

[0109] The processing unit 20023 is used to perform multi-dimensional spatial processing based on multiple contour points of at least one target image region to obtain an initial set of concave and convex point pairs corresponding to each candidate outer contour in at least one target image region.

[0110] Optionally, the processing unit 20023 is specifically used for:

[0111] A multidimensional tree structure is established based on multiple contour points of at least one target image region to obtain multiple pairs of concave and convex points to be screened corresponding to each candidate outer contour in at least one target image region. The first concave and convex point to be screened and the second concave and convex point to be screened in the pair of concave and convex points to be screened are on different sides, and the distance between the first concave and convex point to be screened and the second concave and convex point to be screened is the shortest distance.

[0112] The pairs of concave and convex points that meet the preset concave and convex point pair filtering rules in each candidate outer contour are determined as initial concave and convex point pairs, resulting in multiple initial concave and convex point pairs for each candidate outer contour.

[0113] Multiple initial concave and convex point pairs for each candidate outer contour are combined to obtain at least one set of initial concave and convex point pairs corresponding to each candidate outer contour in the target image region.

[0114] Optionally, the modified filtering module 2004 is specifically used for:

[0115] If there are at least two target initial concave-convex point pairs in the candidate concave-convex point pair set of each candidate outer contour that are in a positional intersection state or meet the preset deletion rules, then at least two target initial concave-convex point pairs are updated to generate at least one target concave-convex point pair corresponding to each candidate outer contour.

[0116] If the number of at least one pair of target concave and convex points for each candidate outer contour is one, then the corresponding candidate outer contour is segmented according to one target concave and convex point to obtain two candidate segmentation regions for each candidate outer contour.

[0117] For each candidate outer contour, the candidate segmentation region that conforms to the preset region rule in the two candidate segmentation regions is determined as the target segmentation region, thus obtaining at least one target segmentation region corresponding to each candidate outer contour.

[0118] If the number of at least two target bump pairs for each candidate outer contour is at least two, then the at least two target bump pairs are classified to obtain the hierarchy of each target bump pair in the at least two target bump pairs.

[0119] Based on the hierarchy of each target concave-convex point pair, the corresponding candidate outer contour is segmented to obtain at least three candidate segmentation regions corresponding to each candidate outer contour.

[0120] Candidate segmentation regions that conform to preset region rules are determined as target segmentation regions in at least three candidate segmentation regions of each candidate outer contour, thereby obtaining at least one target segmentation region corresponding to each candidate outer contour.

[0121] Optionally, map optimization devices may also include:

[0122] The judgment module 2006 is used to determine whether the ratio of the number of contour pixels of each candidate outer contour to the number of corresponding edge obstacle pixels is greater than or equal to a preset proportion if there is no initial concave-convex point pair that meets the preset overlap filtering rules in the initial concave-convex point pair set corresponding to each candidate outer contour.

[0123] The determination module 2007 is used to determine the candidate outer contours with a ratio greater than or equal to a preset proportion as the target outer contours, thereby obtaining at least one target outer contour.

[0124] The AND operation module 2008 is used to perform image filling and dilation processing on at least one target outer contour, and perform AND operation with the initial map to generate a target map.

[0125] Optionally, the map generation module 2005 is specifically used for:

[0126] Image filling is performed on at least one target segmentation region corresponding to each candidate outer contour to obtain at least one filled segmentation region corresponding to each candidate outer contour.

[0127] Dilatation processing is performed on at least one filled segmentation region corresponding to each candidate outer contour to obtain at least one dilated segmentation region corresponding to each candidate outer contour.

[0128] Perform a bitwise AND operation between at least one dilated segmentation region corresponding to each candidate outer contour and the initial map to generate the target map.

[0129] In this embodiment of the invention, image processing is performed on the initial map to obtain at least one initial image region. The at least one initial image region is then subjected to area filtering, multi-dimensional spatial processing, concave-convex point pair correction, and region filtering to obtain at least one target segmentation region. Based on the at least one target segmentation region and the initial map, a target map is generated, thereby deleting reachable areas on the map that the mobile robot cannot reach. This makes the map look more aesthetically pleasing, improves the accuracy of mobile robot positioning and path planning, and enhances the accuracy of the map.

[0130] above Figure 20 and Figure 21 The map optimization device in this embodiment of the invention will be described in detail from the perspective of modular functional entities. The mobile robot in this embodiment of the invention will be described in detail from the perspective of hardware processing.

[0131] Figure 22 This is a schematic diagram of the structure of a mobile robot provided in an embodiment of the present invention. The mobile robot 2200 can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 2210 (e.g., one or more processors) and a memory 2220, and one or more storage media 2230 (e.g., one or more mass storage devices) for storing application programs 22303 or data 22302. The memory 2220 and storage media 2230 can be temporary or persistent storage. The program stored in the storage media 2230 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the mobile robot 2200. Furthermore, the processor 2210 may be configured to communicate with the storage media 2230 and execute the series of instruction operations in the storage media 2230 on the mobile robot 2200.

[0132] The mobile robot 2200 may also include one or more power supplies 2240, one or more wired or wireless network interfaces 2250, one or more input / output interfaces 2260, and / or one or more operating systems 22301, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 22 The illustrated mobile robot structure does not constitute a limitation on the mobile robot and may include more or fewer parts than illustrated, or combine certain parts, or have different arrangements of parts.

[0133] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a mobile robot, cause the mobile robot to perform the steps of the map optimization method.

[0134] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0135] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0136] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A map optimization method, characterized in that, The map optimization method includes: Obtain the initial map to be processed, and perform image processing on the initial map to generate at least one corresponding initial image region, wherein the at least one initial image region includes multiple contour points; Based on the at least one initial image region and the corresponding multiple contour points, area filtering and multi-dimensional spatial processing are performed to obtain at least one target image region and a set of initial concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region. The at least one target image region contains areas that restrict the movement of the mobile robot. If there is at least one target initial concave-convex point pair that meets the preset overlap filtering rule in the initial concave-convex point pair set corresponding to each candidate outer contour, then the at least one target initial concave-convex point pair is added to the candidate concave-convex point pair set of the corresponding candidate outer contour. The preset overlap filtering rule is used to filter out the initial concave-convex point pairs that do not overlap between the two divided regions. For each candidate outer contour, the set of candidate concave and convex point pairs is corrected and the region is filtered to obtain at least one target segmentation region corresponding to each candidate outer contour. The at least one target segmentation region allows the mobile robot to move. A target map is generated based on at least one target segmentation region corresponding to each candidate outer contour and the initial map; The step of performing area filtering and multi-dimensional spatial processing based on the at least one initial image region and the corresponding multiple contour points to obtain at least one target image region and an initial set of concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region includes: calculating the area enclosed by each initial image region in the at least one initial image region; determining the initial image region with an area greater than or equal to a preset area as the target image region to obtain at least one target image region; and performing multi-dimensional spatial processing based on the multiple contour points of the at least one target image region to obtain an initial set of concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region.

2. The map optimization method according to claim 1, characterized in that, The step of acquiring the initial map to be processed and performing image processing on the initial map to generate at least one corresponding initial image region includes: Obtain an initial map to be processed, the initial map including at least one reachable area, at least one obstacle area, and at least one unreachable area; The initial map is binarized to obtain a binarized map; An opening operation is performed on the binarized map to generate at least one corresponding initial image region.

3. The map optimization method according to claim 1, characterized in that, The multi-dimensional spatial processing based on multiple contour points of the at least one target image region yields an initial set of concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region, including: A multidimensional tree structure is established based on multiple contour points of the at least one target image region to obtain multiple pairs of concave and convex points to be screened for each candidate outer contour in the at least one target image region. The first concave and convex point to be screened and the second concave and convex point to be screened in the pair of concave and convex points to be screened are on different sides, and the distance between the first concave and convex point to be screened and the second concave and convex point to be screened is the shortest distance. The pairs of concave and convex points that meet the preset concave and convex point pair filtering rules in each candidate outer contour are determined as initial concave and convex point pairs, resulting in multiple initial concave and convex point pairs for each candidate outer contour. Multiple initial concave and convex point pairs of each candidate outer contour are combined to obtain a set of initial concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region.

4. The map optimization method according to claim 1, characterized in that, The step of performing concave-convex point pair correction and region filtering on the set of candidate concave-convex point pairs corresponding to each candidate outer contour to obtain at least one target segmentation region corresponding to each candidate outer contour includes: If there are at least two target initial concave-convex point pairs in the candidate concave-convex point pair set of each candidate outer contour that are in a positional intersection state or meet the preset deletion rules, then the at least two target initial concave-convex point pairs are updated to generate at least one target concave-convex point pair corresponding to each candidate outer contour. If the number of at least one pair of target concave and convex points for each candidate outer contour is one, then the corresponding candidate outer contour is segmented according to the one target concave and convex point to obtain two candidate segmentation regions for each candidate outer contour. For each candidate outer contour, the candidate segmentation region that conforms to the preset region rule in the two candidate segmentation regions is determined as the target segmentation region, thus obtaining at least one target segmentation region corresponding to each candidate outer contour. If the number of at least two target bump pairs for each candidate outer contour is at least two, then the at least two target bump pairs are classified to obtain the level of each target bump pair in the at least two target bump pairs. Based on the hierarchy of each target concave-convex point pair, the corresponding candidate outer contour is segmented to obtain at least three candidate segmentation regions corresponding to each candidate outer contour. Candidate segmentation regions that conform to the preset region rules in at least three candidate segmentation regions of each candidate outer contour are determined as target segmentation regions, thereby obtaining at least one target segmentation region corresponding to each candidate outer contour.

5. The map optimization method according to any one of claims 1-4, characterized in that, After performing area filtering and multi-dimensional spatial processing based on the at least one initial image region and the corresponding multiple contour points to obtain at least one target image region and an initial set of concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region, the method further includes: If there is no initial concave-convex point pair that meets the preset overlap filtering rule in the initial concave-convex point pair set corresponding to each candidate outer contour, then determine whether the ratio of the number of contour pixels of each candidate outer contour to the number of corresponding edge obstacle pixels is greater than or equal to the preset proportion. Candidate outer contours with a ratio greater than or equal to a preset percentage are identified as target outer contours, thus obtaining at least one target outer contour. The outer contour of the at least one target is filled and dilated using an image, and then ANDed with the initial map to generate a target map.

6. The map optimization method according to claim 1, characterized in that, The step of generating a target map based on at least one target segmentation region corresponding to each candidate outer contour and the initial map includes: Image filling is performed on at least one target segmentation region corresponding to each candidate outer contour to obtain at least one filled segmentation region corresponding to each candidate outer contour. Dilatation processing is performed on at least one filled segmentation region corresponding to each candidate outer contour to obtain at least one dilated segmentation region corresponding to each candidate outer contour. Perform an AND operation between at least one dilated segmentation region corresponding to each candidate outer contour and the initial map to generate the target map.

7. A map optimization device, characterized in that, The map optimization device includes: The acquisition and generation module is used to acquire the initial map to be processed, and perform image processing on the initial map to generate at least one corresponding initial image region, wherein the at least one initial image region includes multiple contour points. The filtering processing module is used to perform area filtering and multi-dimensional spatial processing based on the at least one initial image region and the corresponding multiple contour points to obtain at least one target image region and a set of initial concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region, wherein the at least one target image region contains a region that restricts the movement of the mobile robot. The point pair addition module is used to add the at least one target initial concave-convex point pair to the candidate concave-convex point pair set of the corresponding candidate outer contour if there is at least one target initial concave-convex point pair that meets the preset overlap filtering rules in the initial concave-convex point pair set corresponding to each candidate outer contour. The preset overlap filtering rules are used to filter out the initial concave-convex point pairs that do not overlap between the two divided regions. The correction and filtering module is used to correct the concave and convex point pairs and filter the region for each candidate outer contour corresponding to the set of candidate concave and convex point pairs, so as to obtain at least one target segmentation region corresponding to each candidate outer contour, and the at least one target segmentation region allows the mobile robot to move. The map generation module is used to generate a target map based on at least one target segmentation region corresponding to each candidate outer contour and the initial map; The filtering processing module includes: a calculation unit for calculating the area enclosed by each initial image region in the at least one initial image region; a determination unit for determining the initial image regions with an area greater than or equal to a preset area as target image regions, thereby obtaining at least one target image region; and a processing unit for performing multi-dimensional spatial processing based on multiple contour points of the at least one target image region, thereby obtaining an initial set of concave and convex point pairs corresponding to each candidate outer contour in the at least one target image region.

8. A mobile robot, characterized in that, The mobile robot includes: a memory and at least one processor, wherein the memory stores instructions; The at least one processor invokes the instructions in the memory to cause the mobile robot to execute the map optimization method as described in any one of claims 1-6.

9. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the map optimization method as described in any one of claims 1-6.

Images