Relocation Quality Assessment Method, Apparatus and Computer Storage Medium

Abstract

An embodiment of the present invention discloses a method, apparatus, and computer storage medium for relocalization quality assessment. The method includes: when performing relocalization quality assessment, first obtaining a target local map of the robot's relocalization, and then performing partition matching between the target local map and the map to be matched in the global map to obtain the partition matching scores of each partition. If there are partition matching scores that are less than the preset score threshold and exceed the first preset value, perform a score reduction process on them. In this way, the overall matching score between the map to be matched and the target local map will also decrease, and the quality of this matching can be evaluated more accurately. Furthermore, during relocalization, according to the comprehensive matching scores of each map to be matched in the global map, the maps to be matched that have lower scores and cause interference can be filtered out, improving the accuracy of relocalization.

Description

Technical Field

[0001] The present invention relates to the field of electronic technologies, and in particular, to a method and apparatus for evaluating relocation quality and a computer storage medium. Background Art

[0002] With the rapid development of science and technology, intelligent devices have been widely applied to people's daily lives, bringing great convenience to people's lives. Many movable intelligent devices (such as floor-sweeping robots) face the problem of relocation due to reasons such as artificial movement or device restart. During relocation, the robot usually rotates one week at the starting position, establishes a local probability grid map based on a single-line lidar, and matches it with an existing global map. For each grid on the local map that is matched, different weights are given according to the distance to the nearest point on the global map, and the sum of the weights is calculated to obtain a matching score. On the one hand, since the local map contains less information, good matching results may be obtained in multiple parts of the global map. On the other hand, lidar measurement errors may cause large errors in the local map. Especially in a large-scale indoor environment that is relatively open, it will lead to the local map not obtaining good matching results globally. Therefore, it is necessary to further accurately evaluate the matching quality. Summary of the Invention

[0003] In view of the above problems, the present invention is proposed to provide a method and apparatus for evaluating relocation quality and a computer storage medium that overcome the above problems or at least partially solve the above problems.

[0004] In a first aspect, the present embodiment provides a method for evaluating relocation quality, including:

[0005] Obtaining a target local map of robot relocation;

[0006] After dividing the target local map into N partitions, performing partition matching on the N partitions with a map to be matched in the global map to obtain N partition matching scores corresponding to the N partitions one by one, where N is an integer greater than 1;

[0007] Determining M partition matching scores among the N partition matching scores that are less than a preset score threshold. If M is greater than a first preset value, performing a score reduction process on the N - M partition matching scores among the N partition matching scores except the M partition matching scores to obtain N - M reduced partition matching scores;

[0008] Based on the N - M down - scored partition matching scores and the M partition matching scores, determine the comprehensive matching score between the target local map and the map to be matched. The comprehensive matching score is used to represent the matching degree between the target local map and the map to be matched. The higher the comprehensive matching score, the higher the matching degree between the target local map and the map to be matched.

[0009] Optionally, obtaining the target local map for robot relocalization includes:

[0010] Obtain the initial local map for robot relocalization;

[0011] Scale the initial local map P times at equal intervals within a preset scale range to obtain P local maps, and use each of the P local maps as the target local map in turn.

[0012] Optionally, after dividing the target local map into N partitions and performing partition matching between the N partitions and the map to be matched in the global map to obtain N partition matching scores corresponding one - to - one to the N partitions, it includes:

[0013] Partition the target local map Q times. Each time of partitioning is centered on the current position of the robot. Rotate the N partitions obtained from the previous partitioning around the current position in the same direction by a preset rotation angle to form N partitions corresponding to this partitioning. Among them, the initial N partitions are formed by dividing the target local map centered on the current position at preset central - angle intervals, and the preset rotation angle is not equal to the preset central angle;

[0014] For each partitioning, perform partition matching between the N partitions obtained from this partitioning and the map to be matched to obtain N partition matching scores corresponding one - to - one to the N partitions.

[0015] Optionally, the method further includes:

[0016] Centered on the current position of the robot, traverse the occupied grids in the global map that match the target local map;

[0017] Based on the position information of the occupied grids, correct the comprehensive matching score corresponding to the target local map.

[0018] Optionally, before centering on the current position of the robot and traversing the occupied grids in the global map that match the target local map, the method further includes:

[0019] Obtain P comprehensive matching scores corresponding one - to - one to the P local maps Figure 1 ;

[0020] Determine the top K comprehensive matching scores after sorting the P comprehensive matching scores from approximately small to large;

[0021] Determine whether the comprehensive matching score corresponding to the target local map belongs to the K comprehensive matching scores;

[0022] If so, execute the step: taking the current position of the robot as the center, traverse the occupied grids in the global map that match the target local map.

[0023] Optionally, the method further includes:

[0024] Obtain K corrected comprehensive matching scores corresponding to the K comprehensive matching scores one by one;

[0025] Based on the K corrected comprehensive matching scores, determine the target comprehensive matching score corresponding to the map to be matched.

[0026] Optionally, the correcting the comprehensive matching score corresponding to the target local map based on the position information of the occupied grid includes:

[0027] For each occupied grid, if the number of consecutive occupied grids on the line connecting the center of the occupied grid and the current position of the robot exceeds a second preset value, reduce the matching weight of the consecutive occupied grids, and re-determine the comprehensive matching score corresponding to the target local map.

[0028] In a second aspect, the present embodiment further provides a relocalization quality evaluation device, including:

[0029] An acquisition unit, configured to acquire a target local map for robot relocalization;

[0030] A matching unit, configured to divide the target local map into N partitions and perform partition matching on the N partitions with maps to be matched in the global map to obtain N partition matching scores corresponding to the N partitions one by one, where N is an integer greater than 1;

[0031] A processing unit, configured to determine M partition matching scores among the N partition matching scores that are less than a preset score threshold, and if M is greater than a first preset value, perform a score reduction process on the N - M partition matching scores other than the M partition matching scores among the N partition matching scores to obtain N - M reduced partition matching scores;

[0032] An evaluation unit is configured to determine a comprehensive matching score between the target local map and the map to be matched based on the N-M downscaled partition matching scores and the M partition matching scores. The comprehensive matching score is used to represent the matching degree between the target local map and the map to be matched. The higher the comprehensive matching score, the higher the matching degree between the target local map and the map to be matched.

[0033] Optionally, the obtaining unit is specifically configured to:

[0034] Obtain an initial local map for robot relocalization;

[0035] Perform P equally-spaced scalings on the initial local map within a preset scale range to obtain P local maps, and sequentially use each of the P local maps as the target local map.

[0036] Optionally, the matching unit is specifically configured to:

[0037] Partition the target local map Q times. Each time of partitioning is centered on the current position of the robot. The N partitions obtained from the previous partitioning are rotated around the current position by a preset rotation angle in the same direction to form the N partitions corresponding to the current partitioning. Among them, the initial N partitions are formed by dividing the target local map centered on the current position at preset central angle intervals, and the preset rotation angle is not equal to the preset central angle;

[0038] For each partitioning, perform partition matching between the N partitions obtained from this partitioning and the map to be matched to obtain N partition matching scores corresponding one-to-one to the N partitions.

[0039] Optionally, the apparatus further includes a correction unit, and the correction unit is specifically configured to:

[0040] Centered on the current position of the robot, traverse the occupied grids in the global map that match the target local map;

[0041] Based on the position information of the occupied grids, correct the comprehensive matching score corresponding to the target local map.

[0042] Optionally, the apparatus further includes a judgment unit, and the judgment unit is specifically configured to:

[0043] Before traversing the occupied grids in the global map that match the target local map centered on the current position of the robot, obtain P comprehensive matching scores corresponding one-to-one to the P local maps Figure 1 ;

[0044] Determine the top K comprehensive matching scores after sorting the P comprehensive matching scores from approximately small to large;

[0045] Judge whether the comprehensive matching score corresponding to the target local map belongs to the K comprehensive matching scores;

[0046] If so, execute the step: with the current position of the robot as the center, traverse the occupied grids that match the target local map in the global map.

[0047] Optionally, the evaluation unit is further configured to:

[0048] Obtain K corrected comprehensive matching scores corresponding to the K comprehensive matching scores one by one;

[0049] Based on the K corrected comprehensive matching scores, determine the target comprehensive matching score corresponding to the map to be matched.

[0050] Optionally, the correction unit is specifically configured to:

[0051] For each occupied grid, if the number of consecutive occupied grids on the line connecting the center of the occupied grid and the current position of the robot exceeds a second preset value, reduce the matching weight of the consecutive occupied grids, and re-determine the comprehensive matching score corresponding to the target local map.

[0052] In a third aspect, an embodiment of the present invention further provides a relocalization quality evaluation device, including a processor and a memory: the memory is used to store a program for executing the relocalization quality evaluation method in the foregoing first aspect; the processor is configured to execute the program stored in the memory.

[0053] In a fourth aspect, an embodiment of the present invention further provides a computer storage medium, which is characterized in that it is used to store computer software instructions for the relocalization quality evaluation method in the foregoing first aspect, and it includes a program designed for the adaptive transmission method in the above aspect.

[0054] One or more of the above technical solutions in the embodiments of the present application have at least one or more of the following technical effects:

[0055] In the technical solution of the embodiment of the present invention, when performing relocalization quality assessment, first obtain the target local map of the robot relocalization, and then perform partition matching between the target local map and the map to be matched in the global map to obtain the partition matching scores of each partition. If there are partition matching scores that are less than the preset score threshold and exceed the first preset value, it indicates that the map to be matched and the target local map do not match well in some directions, and the confidence of the matching scores in the remaining directions is not high. Perform a score reduction process on it. In this way, the overall matching score between the map to be matched and the target local map will also decrease, indicating that the matching degree between the map to be matched and the target local map is not high, and the quality of this matching can be evaluated more accurately. Furthermore, during relocalization, according to the comprehensive matching scores of the maps to be matched in the global map, filter out the maps to be matched with lower scores that cause interference, and improve the accuracy of relocalization. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0057] Figure 1 is a flowchart of a relocalization quality assessment method applied to a front-end device in the first embodiment of the present invention;

[0058] Figure 2 is a schematic diagram of a relocalization quality assessment device in the second embodiment of the present invention;

[0059] Figure 3 is a schematic diagram of a relocalization quality assessment device in the third embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] This embodiment discloses a method, apparatus, and computer storage medium for relocalization quality assessment. It can accurately evaluate the matching degree between the map to be matched and the target local map during relocalization to improve the relocalization accuracy. The method includes: obtaining a target local map for robot relocalization; dividing the target local map into N partitions and performing partition matching between the N partitions and the map to be matched in the global map to obtain N partition matching scores corresponding to the N partitions one by one, where N is an integer greater than 1; determining M partition matching scores among the N partition matching scores that are less than a preset score threshold, and if M is greater than a first preset value, performing a score reduction process on the N - M partition matching scores among the N partition matching scores other than the M partition matching scores to obtain N - M reduced partition matching scores; based on the N - M reduced partition matching scores and the M partition matching scores, determining a comprehensive matching score between the target local map and the map to be matched, where the comprehensive matching score is used to represent the matching degree between the target local map and the map to be matched, and the higher the comprehensive matching score, the higher the matching degree between the target local map and the map to be matched.

[0061] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific features in the embodiments of the present application and the embodiments are detailed descriptions of the technical solution of the present application, rather than limitations on the technical solution of the present application. Without conflict, the technical features in the embodiments of the present application and the embodiments can be combined with each other.

[0062] The term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after.

[0063] Embodiment

[0064] The first embodiment of the present invention provides a method for relocalization quality assessment. The flowchart of this method is as Figure 1 shown, including the following steps:

[0065] S101: Obtain a target local map for robot relocalization;

[0066] S102: Divide the target local map into N partitions and perform partition matching between the N partitions and the map to be matched in the global map to obtain N partition matching scores corresponding to the N partitions one by one, where N is an integer greater than 1;

[0067] S103: Determine M partition matching scores among the N partition matching scores that are less than a preset score threshold. If M is greater than a first preset value, perform a score reduction process on the N - M partition matching scores among the N partition matching scores excluding the M partition matching scores to obtain N - M reduced partition matching scores;

[0068] S104: Based on the N - M reduced partition matching scores and the M partition matching scores, determine a comprehensive matching score between the target local map and the map to be matched. The comprehensive matching score is used to represent the matching degree between the target local map and the map to be matched. The higher the comprehensive matching score, the higher the matching degree between the target local map and the map to be matched.

[0069] Specifically, the method in this embodiment can be applied to a moving robot, such as a sweeping robot. Of course, in the specific implementation process, it can also be applied to other mobile devices. Here, the present application does not make any restrictions. This embodiment mainly takes the application to a sweeping robot as an example for a detailed introduction.

[0070] First, in the method of this embodiment, through step S101, when the robot performs re - localization, it rotates one week at the starting position and establishes a local probability grid map based on a single - line lidar, and this map can be directly used as the target local map. Of course, for a more accurate evaluation, step S101 can obtain multiple local maps through a scaling strategy, and each of them is used as the target map to evaluate its matching degree according to the steps in S101 - S104. The specific implementation includes the following steps:

[0071] Obtain the initial local map of the robot's re - localization; perform P equal - interval scalings on the initial local map within a preset scale range to obtain P local maps, and sequentially use each of the P local maps as the target local map.

[0072] Specifically, in this embodiment, after the robot starts and rotates one week, it establishes a local probability grid map based on a single - line lidar and uses this map as the initial local map. Then, according to the given error of the lidar, such as 3%, 5% of the range, etc., with the current position of the robot as the center point, scale the distance from the occupied grid (occupied grid) of the probability grid map to the center point. Taking the given error as 3% of the range as an example, divide the scale into 10 equally - spaced values within the range of - 3% to 3%, and then form 10 new local maps after scaling respectively. In the specific implementation process, the scaling range and interval can be set according to actual needs. Here, the present application does not make any restrictions. After obtaining P local maps through the above - mentioned scaling method, each of the P local maps can be sequentially used as the target local map to obtain the comprehensive matching score corresponding to each local map.

[0073] Furthermore, after determining the target local map, when calculating the comprehensive matching score of the target local map, first, through step S102, the target local map is divided into N partitions, and then the N partitions are subjected to partition matching with the map to be matched in the global map, obtaining N partition matching scores corresponding to the N partitions one by one, where N is an integer greater than 1.

[0074] Specifically, the target local map is matched with the existing global map. For each occupied grid on the target local map that is matched, different weights are given according to the distance to the nearest point on the global map, and the sum of the weights is calculated to obtain the initial matching score. Select several possible regions with higher initial matching scores as the maps to be matched. The maps corresponding to these regions are the maps that the robot may be in during repositioning, and the matching quality of each of them is evaluated in turn according to the method in this embodiment.

[0075] Among them, the global map is the map information saved after the sweeping robot completes a full cleaning, which is the sum of all areas that the robot can reach and detect in a closed environment. The position of the charging pile is also saved in the global map. Both the charging pile and the robot may be moved by the user to another position within the global map, or even may be moved to a place outside the real space corresponding to the global map, such as upstairs and downstairs in a duplex or split-level house type. In this case, the robot needs to perform repositioning. During repositioning, the local map used contains less information, and better matching results may be obtained in multiple local parts of the global map. Moreover, the measurement error of the lidar may cause a large error in the local map, especially in a large-scale indoor environment that is relatively empty, which will lead to the local map not getting a good matching result globally. Therefore, it is necessary to further evaluate the matching quality.

[0076] When evaluating the matching quality, the feature matching method is usually adopted. In an indoor environment, the wall line is a very strong feature. Similar wall lines may exist in different corners of the global map, thus interfering with the matching result. Considering the regional distribution of the wall lines, they often distribute in one or two directions of the local map, and the matching results in the remaining directions actually have discrimination. Therefore, the method in this embodiment adopts the partition matching evaluation method. Specifically, first, the target local map is divided into N partitions, and then partition matching is performed on the maps to be matched obtained in the foregoing embodiment, that is, the sum of the weights of the occupied grids corresponding to each partition is calculated to obtain the partition matching score corresponding to each partition in the N partitions.

[0077] Specifically, when partitioning the target local map, it can be achieved through the following steps:

[0078] Perform Q partitions on the target local map. Each partition is centered around the current position of the robot. The N partitions obtained from the previous partition are rotated around the current position in the same direction by a preset rotation angle to form the N partitions corresponding to this partition. Among them, the initial N partitions are formed by dividing the target local map centered around the current position at preset central angle intervals, and the preset rotation angle is not equal to the preset central angle. For each partition, perform partition matching between the N partitions obtained from this partition and the map to be matched, and obtain N partition matching scores corresponding one-to-one to the N partitions.

[0079] Specifically, in this embodiment, when partitioning, multiple rays can be generated centered around the current position of the robot at the same preset central angle (such as 60 degrees, 90 degrees, etc.). The area enclosed by two rays is a partition. In the specific implementation process, the number of divided areas can be set according to actual needs. For example: 4 partitions, 6 partitions, etc. In specific applications, houses are usually square, and wall line features are distributed in 4 directions. Therefore, this embodiment mainly introduces the case of 4 partitions in detail.

[0080] Emit 4 rays from the center, with an interval of 90 degrees between adjacent rays, dividing the local map evenly into 4 regions. Select an appropriate preset rotation angle (such as: 20 degrees, 60 degrees, etc.) to perform counterclockwise circular rotation on the 4 regions around the center in the same direction. Each time it rotates, a new partition can be formed. Each partition is divided into 4 regions until it rotates one full circle, and Q partitions are performed. The distribution of occupied grids that match the target local map in the entire space is the same. The partition actually divides the space into 4 regions each occupying a 90° direction centered around the current position of the robot. For a good match, geometrically speaking, if the structure of the real space has not changed significantly, relatively high matching scores should be obtained in all directions. Rotating multiple times and performing multiple partitions is an optimization search process to obtain those directions with poor matches as much as possible to make the evaluation result of the matching accuracy more accurate. In the specific implementation process, the partition method and the number of partitions can be set according to actual needs, and this application does not limit them here.

[0081] Furthermore, for each partition resulting in N partitions, the partition matching score of each partition will be calculated. Taking the 4 partitions in the above embodiment as an example, after each partition forms 4 partitions, the partition matching score corresponding to each partition is obtained according to the aforementioned weight sum calculation method.

[0082] In this way, the partition matching scores corresponding to each of the N partitions formed each time can be obtained. Then, through step S103, M partition matching scores less than the preset score threshold among the N partition matching scores are determined. If M is greater than the first preset value, the N - M partition matching scores other than the M partition matching scores among the N partition matching scores are subject to a score reduction process to obtain N - M partition matching scores after the score reduction.

[0083] Specifically, continuing with the above example of 4 partitions, 4 partitions are obtained each time of partitioning, assumed to be A, B, C, and D. Each partition corresponds to a partition matching score. Assume A = 0.2, B = 0.2, C = 0.8, D = 0.8. If calculated according to the existing technology, the score of this match is (A + B + C + D) / 4 = 0.5. This value is in the middle section of 0 - 1, and it is difficult to judge the matching degree between the map to be matched and the target local map based on this. Therefore, further evaluation is needed. It is found that the partition matching scores in regions A and B are both relatively low, both less than the score threshold of 0.4, indicating that there is a low matching degree in the two directions of A and B, and the existing region is greater than the first preset value of 1. Then, for the relatively well - matched regions C and D, their matching confidence will also decrease. Therefore, the partition matching scores of partitions C and D are subject to a score reduction process. There are various ways to implement the score reduction process. For example: a relatively low preset weight coefficient of 0.2 can be assigned to C and D. In this way, the partition matching scores of partitions C and D after the score reduction are both 0.2 * 0.8 = 0.16. Another example: a preset subtraction value of 0.4 is assigned to C and D. In this way, the partition matching scores of partitions C and D after the score reduction are both 0.8 - 0.4 = 0.4. In the specific implementation process, the first preset value, the preset score threshold, and the way of the score reduction process can all be set according to actual needs. Here, the present application does not make any restrictions.

[0084] Furthermore, after obtaining the N - M partition matching scores and the M partition matching scores after the score reduction process, the comprehensive matching score between the target local map and the map to be matched can be obtained through step S104. This comprehensive matching score is used to indicate the quality of this matching evaluation. The larger this value is, the better the matching; the smaller this value is, the lower the matching degree. Continuing with the previous example, A = 0.2, B = 0.2, and C = D = 0.16 after the score reduction process. The comprehensive matching score is (0.2 + 0.0 + 0.16 + 0.16) / 4 = 0.18. This indicates that the quality of this match is not high, and the map to be matched can be excluded.

[0085] In another case, assume A = 0.5, B = 0.5, C = 0.5, D = 0.5. If calculated according to the existing technology, the score of this match is (A + B + C + D) / 4 = 0.5. This score is in the middle of 0 - 1, and it is very difficult to judge the matching degree between the map to be matched and the target local map based on this. Therefore, further evaluation is required. It is found that the partition matching scores in areas A, B, C, and D are very average, all greater than the score threshold of 0.4, indicating that A, B, C, and D are evenly matched in 4 directions, and the confidence level of their matching is relatively high. No processing is required, and the comprehensive matching score is 0.5. The map to be matched can be used as an alternative map for further position determination in the subsequent process.

[0086] As can be seen from the foregoing embodiments, if there are P target local maps obtained by using the scaling strategy, P target local maps can be obtained through steps S102 - S104. Figure 1 The corresponding P comprehensive matching scores. Here, for the specific process, refer to the foregoing embodiments, and this application will not elaborate here. For these P comprehensive matching scores, they can be sorted from large to small, and the maximum value is used as the final matching score of the map to be matched.

[0087] Qualitatively speaking, each room, or even different rooms, may be very similar at the 4 corners, because there are two wall lines at the corner in the map, resulting in a falsely high score for global matching, and actually having no discrimination. What the method in this embodiment does is to try to find the directions with poor matching as much as possible, because these directions have discrimination. If the partition scores in the first preset number of directions are relatively small, it means that the result of this global matching is not very credible. Therefore, the scores of the other partitions with high scores are reduced, and then the global weighted average is calculated, and the comprehensive matching score will be reduced numerically, indicating that the matching degree between the map to be matched and the target local map is not high, and the quality of this match can be evaluated more accurately. Furthermore, during re - positioning, according to the comprehensive matching scores of each map to be matched in the global map, the maps to be matched with lower scores that cause interference can be filtered out, improving the accuracy of re - positioning.

[0088] Further, for two adjacent rooms, when performing global re - positioning, the wall line in one direction may be matched to the wall of another room, resulting in an incorrect re - positioning result. To more accurately evaluate the matching accuracy, the method in this embodiment will also perform wall - penetration correction on the map to be matched. The wall - penetration correction can be achieved through the following steps:

[0089] Taking the current position of the robot as the center, traverse the occupied grids in the global map that match the target local map, and based on the position information of the occupied grids, correct the comprehensive matching score corresponding to the target local map.

[0090] Among them, if there are P target local maps obtained by using the scaling strategy, the following steps are further included: obtaining P comprehensive matching scores corresponding to the P local maps one by one; sorting the P comprehensive matching scores from small to large and determining the top K comprehensive matching scores; judging whether the comprehensive matching score corresponding to the target local map belongs to the K comprehensive matching scores; if so, executing the step of traversing the occupied grids matching the target local map in the global map with the current position of the robot as the center. Figure 1 Furthermore, based on the position information of the occupied grids, the comprehensive matching score corresponding to the target local map is corrected, including: for each occupied grid, if the number of consecutive occupied grids on the line connecting the center of the occupied grid and the current position of the robot exceeds a second preset value, reducing the matching weight of the consecutive occupied grids, and re-determining the comprehensive matching score corresponding to the target local map.

[0091] Specifically, if there is only one target local map, referring to the case where the local map scanned by the robot rotating one circle at startup is used as the target map mentioned in the foregoing embodiment, the wall-penetration correction is directly performed on it. If there are P target local maps obtained by using the scaling strategy, the local maps corresponding to the top K comprehensive matching scores sorted from small to large in score value are selected from the P comprehensive matching scores for wall-penetration correction.

[0092] When performing wall-penetration correction, for the target local map, with the current position of the robot as the center, traverse the occupied grids (occupied grids) matching the target local map in the global map, take the center point of each grid, and connect it with the current position point of the robot to form a line segment respectively. Check the grids passed by the line segment. If more than the second preset value of consecutive n grids are in the occupied state, it is considered that the n grids penetrate the wall, and the matching weight of the n grids is reduced. After the traversal is completed, calculate the sum of the weights of all grids as the matching score.

[0093] In the specific implementation process, the size of the second preset value is related to the house type structure and the map resolution. For example, the length and width of a grid on the map are 5 cm. Assuming that the wall thickness of a common house type is greater than 20 cm, that is, the wall thickness occupies more than 5 grids, then the second preset value can be selected as 5, and a balanced value can be given according to the data of different house types and the wall-penetration angle. In the specific implementation process, the second preset value can be set according to actual needs, and this application does not make any restrictions here.

[0094] Furthermore, if there is only one target local map, the comprehensive matching score after wall-penetration correction can be used as the final target comprehensive matching score for the map to be matched, and the matching quality of this time can be determined accordingly.

[0095]

[0096] ​When there are P target local maps, P corresponding comprehensive matching scores are obtained according to the steps of S102 - S104. After selecting the K local maps corresponding to the K larger comprehensive matching scores for wall - penetration correction, K corrected comprehensive matching scores corresponding one - to - one with the K comprehensive matching scores are obtained. Based on the K corrected comprehensive matching scores, the target comprehensive matching score corresponding to the map to be matched is determined. Specifically, the maximum value among the K corrected comprehensive matching scores can be selected as the final target comprehensive matching score for this map to be matched, and the matching quality of this time is judged accordingly. Of course, the average value of the K corrected comprehensive matching scores can also be taken as the final target comprehensive matching score for this map to be matched, and the matching quality of this time is judged accordingly. In the specific implementation process, the value of K can be set according to actual needs, and here, this application does not make any restrictions.

[0097] The solution in this embodiment adopts a scaling strategy to generate P local maps, then performs partition matching on them, obtains comprehensive matching scores according to the partition matching scores of each partition, then selects the ones with larger scores for wall - penetration correction, and takes the largest comprehensive matching score after correction as the evaluation score of this matching. It can have better robustness and higher accuracy.

[0098] Please refer to Figure 2 , the second embodiment of the present invention also provides a relocalization quality evaluation device, and the device includes:

[0099] An acquisition unit 201, configured to acquire the target local map for robot relocalization;

[0100] A matching unit 202, configured to divide the target local map into N partitions and perform partition matching on the N partitions with the map to be matched in the global map, to obtain N partition matching scores corresponding one - to - one to the N partitions, where N is an integer greater than 1;

[0101] A processing unit 203, configured to determine M partition matching scores among the N partition matching scores that are less than a preset score threshold. If M is greater than a first preset value, perform a score - reduction process on the N - M partition matching scores among the N partition matching scores except the M partition matching scores, to obtain N - M reduced partition matching scores;

[0102] An evaluation unit 204, configured to determine the comprehensive matching score between the target local map and the map to be matched based on the N - M reduced partition matching scores and the M partition matching scores. The comprehensive matching score is used to represent the matching degree between the target local map and the map to be matched. The higher the comprehensive matching score, the higher the matching degree between the target local map and the map to be matched.

[0103] As an alternative embodiment, the obtaining unit 201 is specifically configured to:

[0104] Obtain an initial local map for robot relocalization;

[0105] Perform P equally-spaced scalings on the initial local map within a preset scale range to obtain P local maps, and sequentially use each of the P local maps as a target local map.

[0106] As an alternative embodiment, the matching unit 202 is specifically configured to:

[0107] Partition the target local map Q times. Each partition is centered on the current position of the robot. The N partitions obtained from the previous partition are rotated around the current position in the same direction by a preset rotation angle to form the N partitions corresponding to the current partition. Among them, the initial N partitions are formed by dividing the target local map at preset central angle intervals centered on the current position, and the preset rotation angle is not equal to the preset central angle;

[0108] For each partition, perform partition matching between the N partitions obtained from this partition and the map to be matched to obtain N partition matching scores corresponding one-to-one to the N partitions.

[0109] As an alternative embodiment, the apparatus further includes a correction unit, and the correction unit is specifically configured to:

[0110] Traverse the occupancy grids in the global map that match the target local map centered on the current position of the robot;

[0111] Based on the position information of the occupancy grids, correct the comprehensive matching score corresponding to the target local map.

[0112] As an alternative embodiment, the apparatus further includes a judgment unit, and the judgment unit is specifically configured to:

[0113] Before traversing the occupancy grids in the global map that match the target local map centered on the current position of the robot, obtain P comprehensive matching scores corresponding one-to-one to the P local maps Figure 1 Sort the P comprehensive matching scores from approximately small to large and determine the top K comprehensive matching scores;

[0114] Judge whether the comprehensive matching score corresponding to the target local map belongs to the K comprehensive matching scores;

[0115]

[0116] ​If so, execute the steps: with the current position of the robot as the center, traverse the occupied grids in the global map that match the target local map.

[0117] As an optional embodiment, the evaluation unit 204 is further configured to:

[0118] Obtain K corrected comprehensive matching scores corresponding to the K comprehensive matching scores one by one;

[0119] Based on the K corrected comprehensive matching scores, determine the target comprehensive matching score corresponding to the map to be matched.

[0120] As an optional embodiment, the correction unit is specifically configured to:

[0121] For each occupied grid, if the number of consecutive occupied grids on the line connecting the center of the occupied grid and the current position of the robot exceeds a second preset value, reduce the matching weight of the consecutive occupied grids, and re-determine the comprehensive matching score corresponding to the target local map.

[0122] Specifically, in this embodiment, the specific implementation manner of data transmission by the relocalization quality evaluation device has been elaborated in detail in the foregoing first embodiment, and will not be repeated here.

[0123] The third embodiment of the present invention further provides a relocalization quality evaluation device. Please refer to Figure 3 , for ease of explanation, only the parts related to the embodiments of the present invention are shown. For the specific technical details not disclosed, please refer to the method part of the embodiments of the present invention.

[0124] Figure 3 Shown is a schematic diagram of a part of the structure of the relocalization quality evaluation device provided by the embodiments of the present invention. The relocalization quality evaluation device includes a memory 301, and the memory 301 is used to store a program for executing the relocalization quality evaluation method in the foregoing first embodiment. The relocalization quality evaluation device further includes a processor 302, connected to the memory 301, and the processor 302 is configured to execute the program stored in the memory 301.

[0125] When the processor 302 executes the computer program, it implements the steps in the relocalization quality evaluation method in the foregoing first embodiment. Alternatively, when the processor executes the computer program, it implements the functions of each module / unit in the relocalization quality evaluation device in the foregoing second embodiment.

[0126] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing specific functions, and these instruction segments are used to describe the execution process of the computer program in the computer device.

[0127] The device may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that the schematic Figure 3 is merely an example diagram of the functional components of the relocation quality assessment device, and does not constitute a limitation on the relocation quality assessment device. It may include more or fewer components than those shown in the figure, or combine certain components, or different components. For example, the relocation quality assessment device may further include input / output devices, network access devices, buses, etc.

[0128] The so-called processor 302 may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the computer device, and connects various parts of the entire computer device through various interfaces and lines.

[0129] The memory 301 can be used to store the computer program and / or module. By running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory, the processor realizes various functions of the computer device. The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created according to the use of the relocalization quality assessment device (such as audio data, video data, etc.). In addition, the memory can include high-speed random access memory, and can also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0130] In an embodiment of the present invention, the processor 302 has the following functions:

[0131] Obtain the target local map for robot relocalization;

[0132] After dividing the target local map into N partitions, perform partition matching on the N partitions with the maps to be matched in the global map to obtain N partition matching scores corresponding to the N partitions one by one, where N is an integer greater than 1;

[0133] Determine M partition matching scores among the N partition matching scores that are less than a preset score threshold. If M is greater than a first preset value, perform a score reduction process on the N - M partition matching scores other than the M partition matching scores among the N partition matching scores to obtain N - M reduced partition matching scores;

[0134] Based on the N - M reduced partition matching scores and the M partition matching scores, determine the comprehensive matching score between the target local map and the map to be matched. The comprehensive matching score is used to represent the matching degree between the target local map and the map to be matched. The higher the comprehensive matching score, the higher the matching degree between the target local map and the map to be matched.

[0135] In an embodiment of the present invention, the processor 302 also has the following functions:

[0136] Obtain the initial local map for robot relocalization;

[0137] Scale the initial local map P times at equal intervals within a preset scale range to obtain P local maps, and sequentially use each of the P local maps as the target local map.

[0138] In an embodiment of the present invention, the processor 302 further has the following functions:

[0139] Partition the target local map Q times. Each time of partitioning is centered on the current position of the robot. Rotate the N partitions obtained from the previous partitioning around the current position in the same direction by a preset rotation angle to form N partitions corresponding to the current partitioning. Among them, the initial N partitions are formed by dividing the target local map centered on the current position at preset central angle intervals, and the preset rotation angle is not equal to the preset central angle;

[0140] For each partitioning, perform partitioning matching between the N partitions obtained from this partitioning and the map to be matched to obtain N partitioning matching scores corresponding one by one to the N partitions.

[0141] In an embodiment of the present invention, the processor 302 further has the following functions:

[0142] Centered on the current position of the robot, traverse the occupied grids in the global map that match the target local map;

[0143] Based on the position information of the occupied grids, correct the comprehensive matching score corresponding to the target local map.

[0144] In an embodiment of the present invention, the processor 302 further has the following functions:

[0145] Before traversing the occupied grids in the global map that match the target local map centered on the current position of the robot, obtain P comprehensive matching scores corresponding one by one to the P local maps Figure 1 ;

[0146] Sort the P comprehensive matching scores from small to large and determine the top K comprehensive matching scores;

[0147] Judge whether the comprehensive matching score corresponding to the target local map belongs to the K comprehensive matching scores;

[0148] If so, execute the step: Centered on the current position of the robot, traverse the occupied grids in the global map that match the target local map.

[0149] In an embodiment of the present invention, the processor 302 further has the following functions:

[0150] Obtain K corrected comprehensive matching scores corresponding to the K comprehensive matching scores one by one;

[0151] Based on the K corrected comprehensive matching scores, determine a target comprehensive matching score corresponding to the map to be matched.

[0152] In an embodiment of the present invention, the processor 302 further has the following functions:

[0153] For each occupied grid, if the number of consecutive occupied grids existing on the line connecting the center of the occupied grid and the current position of the robot exceeds a second preset value, reduce the matching weight of the consecutive occupied grids, and re-determine the comprehensive matching score corresponding to the target local map.

[0154] A fourth embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. If the functional units integrated in the relocalization quality evaluation device in the second embodiment of the present invention are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, all or part of the processes in the relocalization quality evaluation method of the first embodiment of the present invention can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device, medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0155] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they know the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present invention.

[0156] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

[0157] The present invention also discloses A1, a method for evaluating the quality of relocalization, including:

[0158] Obtaining a target local map of the robot relocalization;

[0159] After dividing the target local map into N partitions, performing partition matching between the N partitions and the map to be matched in the global map to obtain N partition matching scores corresponding to the N partitions one by one, where N is an integer greater than 1;

[0160] Determining M partition matching scores less than a preset score threshold among the N partition matching scores. If M is greater than a first preset value, performing a score reduction process on the N - M partition matching scores other than the M partition matching scores among the N partition matching scores to obtain N - M reduced partition matching scores;

[0161] Based on the N - M reduced partition matching scores and the M partition matching scores, determining a comprehensive matching score between the target local map and the map to be matched, where the comprehensive matching score is used to represent the matching degree between the target local map and the map to be matched, and the higher the comprehensive matching score, the higher the matching degree between the target local map and the map to be matched.

[0162] A2. The method according to A1, wherein the obtaining of the target local map of the robot relocalization includes:

[0163] Obtaining an initial local map of the robot relocalization;

[0164] Performing P equal - interval scalings on the initial local map within a preset scale range to obtain P local maps, and sequentially using each of the P local maps as the target local map.

[0165] A3. The method according to A1, wherein the performing partition matching between the N partitions and the map to be matched in the global map after dividing the target local map into N partitions to obtain N partition matching scores corresponding to the N partitions one by one includes:

[0166] Perform Q partitions on the target local map. Each partition takes the current position of the robot as the center, and rotates the N partitions obtained from the previous partition around the current position in the same direction by a preset rotation angle to form the N partitions corresponding to the current partition. Among them, the initial N partitions are formed by dividing the target local map centered on the current position at preset central angle intervals, and the preset rotation angle is not equal to the preset central angle;

[0167] For each partition, perform partition matching on the N partitions obtained from this partition and the map to be matched, and obtain N partition matching scores corresponding one by one to the N partitions.

[0168] A4. The method as described in A2, the method further includes:

[0169] Centered on the current position of the robot, traverse the occupied grids in the global map that match the target local map;

[0170] Based on the position information of the occupied grids, correct the comprehensive matching score corresponding to the target local map.

[0171] A5. The method as described in A2, before traversing the occupied grids in the global map that match the target local map centered on the current position of the robot, the method further includes:

[0172] Obtain P comprehensive matching scores corresponding one by one to the P local maps Figure 1 ;

[0173] Sort the P comprehensive matching scores from smallest to largest and determine the top K comprehensive matching scores;

[0174] Judge whether the comprehensive matching score corresponding to the target local map belongs to the K comprehensive matching scores;

[0175] If so, execute the step: centered on the current position of the robot, traverse the occupied grids in the global map that match the target local map.

[0176] A6. The method as described in A5, the method further includes:

[0177] Obtain K corrected comprehensive matching scores corresponding one by one to the K comprehensive matching scores;

[0178] Based on the K corrected comprehensive matching scores, determine the target comprehensive matching score corresponding to the map to be matched.

[0179] A7. The method as described in A4, the correcting the comprehensive matching score corresponding to the target local map based on the position information of the occupied grids includes:

[0180] For each occupied grid, if the number of consecutive occupied grids on the line connecting the center of the occupied grid and the current position of the robot exceeds a second preset value, reduce the matching weight of the consecutive occupied grids, and re-determine the comprehensive matching score corresponding to the target local map.

[0181] B8. A relocalization quality assessment device, comprising:

[0182] An acquisition unit, configured to acquire a target local map of robot relocalization;

[0183] A matching unit, configured to divide the target local map into N partitions and perform partition matching on the N partitions with a map to be matched in a global map to obtain N partition matching scores corresponding to the N partitions one by one, where N is an integer greater than 1;

[0184] A processing unit, configured to determine M partition matching scores less than a preset score threshold among the N partition matching scores. If M is greater than a first preset value, perform a score reduction process on the N - M partition matching scores other than the M partition matching scores among the N partition matching scores to obtain N - M reduced partition matching scores;

[0185] An evaluation unit, configured to determine a comprehensive matching score of the target local map and the map to be matched based on the N - M reduced partition matching scores and the M partition matching scores. The comprehensive matching score is used to represent the matching degree between the target local map and the map to be matched. The higher the comprehensive matching score, the higher the matching degree between the target local map and the map to be matched.

[0186] B9. The device according to B8, wherein the acquisition unit is specifically configured to:

[0187] Obtain an initial local map of robot relocalization;

[0188] Perform P equal - interval scalings on the initial local map within a preset scale range to obtain P local maps, and sequentially use each of the P local maps as the target local map.

[0189] B10. The device according to B8, wherein the matching unit is specifically configured to:

[0190] Partition the target local map Q times. Each time of partitioning takes the current position of the robot as the center, and rotates the N partitions obtained from the previous partitioning around the current position in the same direction by a preset rotation angle to form the N partitions corresponding to the current partitioning. Among them, the initial N partitions are formed by dividing the target local map centered on the current position at preset central angle intervals, and the preset rotation angle is not equal to the preset central angle;

[0191] For each partitioning, perform partitioning matching on the N partitions obtained from this partitioning and the map to be matched to obtain N partition matching scores corresponding one by one to the N partitions.

[0192] B11. The device as described in B9, the device further includes a correction unit, and the correction unit is specifically used for:

[0193] Taking the current position of the robot as the center, traverse the occupied grids in the global map that match the target local map;

[0194] Based on the position information of the occupied grids, correct the comprehensive matching score corresponding to the target local map.

[0195] B12. The device as described in B9, the device further includes a judgment unit, and the judgment unit is specifically used for:

[0196] Before traversing the occupied grids in the global map that match the target local map with the current position of the robot as the center, obtain P comprehensive matching scores corresponding one by one to the P local maps; Figure 1 A corresponding P comprehensive matching scores;

[0197] Sort the P comprehensive matching scores from approximately small to large and determine the top K comprehensive matching scores;

[0198] Judge whether the comprehensive matching score corresponding to the target local map belongs to the K comprehensive matching scores;

[0199] If so, execute the step: taking the current position of the robot as the center, traverse the occupied grids in the global map that match the target local map.

[0200] B13. The device as described in B12, the evaluation unit is further used for:

[0201] Obtain K corrected comprehensive matching scores corresponding one by one to the K comprehensive matching scores;

[0202] Based on the K corrected comprehensive matching scores, determine the target comprehensive matching score corresponding to the map to be matched.

[0203] B14. The device according to B11, wherein the calibration unit is specifically configured to:

[0204] For each occupied grid, if the number of consecutive occupied grids existing on the line connecting the center of the occupied grid and the current position of the robot exceeds a second preset value, reduce the matching weight of the consecutive occupied grids, and re-determine the comprehensive matching score corresponding to the target local map.

[0205] C15. A relocalization quality evaluation device, comprising a processor and a memory:

[0206] The memory is used to store a program for executing the method according to any one of A1 to A7;

[0207] The processor is configured to execute the program stored in the memory.

[0208] D16. A computer storage medium for storing computer software instructions for the relocalization quality evaluation method according to any one of A1 to A7 above, which includes a program designed for the adaptive transmission method in the above aspects.

Claims

1. A relocation quality assessment method, characterized in that Including: Obtaining a target local map for robot relocalization; Dividing the target local map into N partitions, and then performing partition matching between the N partitions and a map to be matched in the global map to obtain N partition matching scores corresponding to the N partitions one by one, where N is an integer greater than 1; Determining M partition matching scores that are less than a preset score threshold among the N partition matching scores. If M is greater than a first preset value, performing a score reduction process on the N - M partition matching scores other than the M partition matching scores among the N partition matching scores to obtain N - M partition matching scores after score reduction; Based on the N - M partition matching scores after score reduction and the M partition matching scores, determining a comprehensive matching score between the target local map and the map to be matched. The comprehensive matching score is used to represent the matching degree between the target local map and the map to be matched. The higher the comprehensive matching score, the higher the matching degree between the target local map and the map to be matched.

2. The method according to claim 1, wherein The obtaining of the target local map for robot relocalization includes: Obtaining an initial local map for robot relocalization; Performing P equal - interval scalings on the initial local map within a preset scale range to obtain P local maps, and sequentially using each of the P local maps as the target local map.

3. The method according to claim 1, characterized in that The dividing the target local map into N partitions and then performing partition matching between the N partitions and the map to be matched in the global map to obtain N partition matching scores corresponding to the N partitions one by one includes: Performing Q partitions on the target local map. Each partition takes the current position of the robot as the center, and rotates the N partitions obtained from the previous partition around the current position in the same direction by a preset rotation angle to form N partitions corresponding to the current partition. Among them, the initial N partitions are formed by dividing the target local map at preset central - angle intervals with the current position as the center, and the preset rotation angle is not equal to the preset central angle; For each partition, performing partition matching between the N partitions obtained from this partition and the map to be matched to obtain N partition matching scores corresponding to the N partitions one by one.

4. The method according to claim 2, wherein The method further includes: Taking the current position of the robot as the center, traversing the occupancy grids in the global map that match the target local map; Based on the position information of the occupancy grids, correcting the comprehensive matching score corresponding to the target local map.

5. The method according to claim 4, wherein Before taking the current position of the robot as the center and traversing the occupancy grids in the global map that match the target local map, the method further includes: Obtaining P comprehensive matching scores corresponding to the P local maps one by one; Sorting the P comprehensive matching scores from smallest to largest and determining the top K comprehensive matching scores; Judging whether the comprehensive matching score corresponding to the target local map belongs to the K comprehensive matching scores; If so, performing the step: taking the current position of the robot as the center, traversing the occupancy grids in the global map that match the target local map.

6. The method according to claim 5, wherein The method further includes: Obtain K corrected comprehensive matching scores corresponding to the K comprehensive matching scores one by one; Based on the K corrected comprehensive matching scores, determine the target comprehensive matching score corresponding to the map to be matched.

7. The method according to claim 4, wherein The correcting the comprehensive matching score corresponding to the target local map based on the position information of the occupied grid includes: For each occupied grid, if the number of consecutive occupied grids on the line connecting the center of the occupied grid and the current position of the robot exceeds a second preset value, reduce the matching weight of the consecutive occupied grids, and re-determine the comprehensive matching score corresponding to the target local map.

8. A relocation quality assessment device, characterized in that, Including: An acquisition unit, configured to acquire a target local map for robot relocalization; A matching unit, configured to divide the target local map into N partitions and perform partition matching on the N partitions with a map to be matched in the global map to obtain N partition matching scores corresponding to the N partitions one by one, where N is an integer greater than 1; A processing unit, configured to determine M partition matching scores among the N partition matching scores that are less than a preset score threshold. If M is greater than a first preset value, perform a score reduction process on the N - M partition matching scores other than the M partition matching scores among the N partition matching scores to obtain N - M reduced partition matching scores; An evaluation unit, configured to determine the comprehensive matching score between the target local map and the map to be matched based on the N - M reduced partition matching scores and the M partition matching scores. The comprehensive matching score is used to represent the matching degree between the target local map and the map to be matched. The higher the comprehensive matching score, the higher the matching degree between the target local map and the map to be matched.

9. The device according to claim 8, wherein, The acquisition unit is specifically configured to: Obtain an initial local map for robot relocalization; Perform P equal-interval scalings on the initial local map within a preset scale range to obtain P local maps, and use each of the P local maps as the target local map in turn.

10. The device according to claim 8, characterized in that, The matching unit is specifically configured to: Perform Q partitions on the target local map. Each partition is centered on the current position of the robot. The N partitions obtained from the previous partition are rotated around the current position in the same direction by a preset rotation angle to form N partitions corresponding to the current partition. Among them, the initial N partitions are formed by dividing the target local map at preset central angle intervals centered on the current position, and the preset rotation angle is not equal to the preset central angle; For each partition, perform partition matching on the N partitions obtained from this partition with the map to be matched to obtain N partition matching scores corresponding to the N partitions one by one.

11. The device according to claim 9, characterized in that The apparatus further includes a correction unit, and the correction unit is specifically configured to: Centered on the current position of the robot, traverse the occupied grids in the global map that match the target local map; Based on the position information of the occupied grid, correct the comprehensive matching score corresponding to the target local map.

12. The device according to claim 11, wherein The apparatus further includes a judgment unit, and the judgment unit is specifically configured to: Before traversing the occupancy grids in the global map that match the target local map centered on the current position of the robot, obtain P comprehensive matching scores corresponding one by one to the P local maps; After sorting the P comprehensive matching scores from roughly small to large, determine the top K comprehensive matching scores; Judge whether the comprehensive matching score corresponding to the target local map belongs to the K comprehensive matching scores; If so, execute the step: Traverse the occupancy grids in the global map that match the target local map centered on the current position of the robot.

13. The device according to claim 12, characterized in that, The evaluation unit is further configured to: Obtain K corrected comprehensive matching scores corresponding one by one to the K comprehensive matching scores; Based on the K corrected comprehensive matching scores, determine the target comprehensive matching score corresponding to the map to be matched.

14. The device according to claim 11, wherein The correction unit is specifically configured to: For each occupancy grid, if the number of consecutive occupancy grids existing on the line connecting the center of the occupancy grid and the current position of the robot exceeds a second preset value, reduce the matching weight of the consecutive occupancy grids, and re-determine the comprehensive matching score corresponding to the target local map.

15. A relocation quality assessment device, characterized in that, Comprising a processor and a memory: The memory is used to store a program for executing the method according to any one of claims 1 to 7; The processor is configured to execute the program stored in the memory.

16. A computer storage medium, characterized in that, For storing computer software instructions for the relocalization quality evaluation method according to any one of the above claims 1 to 7, which include programs designed for the relocalization quality evaluation method.

Images