A high-precision map updating method and system

Abstract

This invention provides a high-precision map updating method and system. The method includes: performing image registration and fixed control point registration on map fragments based on multiple sampling trajectories; segmenting and clustering the elements based on the center points of each reference master map; establishing a quality function based on the confidence values ​​and spatial relationships of the elements in each map fragment; and integrating the quality function and calculating the confidence to update the map elements using the DS evidence theory. This method determines the positional relationship of the elements in the map fragment with other samples based on their own position, providing important guidance for improving the accuracy and efficiency of geometric shape recognition.

Description

Technical Field

[0001] The present invention relates to the field of computer vision and mapping, and more specifically, to a high-precision map updating method and system. Background Art

[0002] The fusion of relative position confidence values ​​for geometric shapes utilizes a large amount of input local fragment map data to combine the information covered by the confidence values ​​of each established ground feature into an overall evaluation confidence value. This update uses the fused overall evaluation index to determine changes in map ground features and then decide whether to add, remove, or modify ground features on the current high-precision map, thereby maximizing the timeliness and accuracy of the high-precision map. Summary of the Invention

[0003] The present invention aims to solve the technical problems existing in the prior art and provides a high-precision map updating method and system.

[0004] According to a first aspect of the present invention, a high-precision map updating method is provided, comprising:

[0005] Perform ground feature registration based on the fragmented maps sampled multiple times and the reference master map to form a set of map feature registration pairs;

[0006] Perform concentric circle pixel segmentation and clustering based on the center points of each element in the reference mother map, calculate the total distance loss between each pair of fragment map elements and the mother map elements in the map element registration pair set, and tentatively assign a status label to each fragment map element based on the total distance loss. The status label includes whether the newly added map element, the deleted map element, and the map element corresponding to the current mother map are of the same category.

[0007] Establish a corresponding quality function based on the confidence value of each fragment map element;

[0008] Calculating the credibility function of each state of each fragment map element based on the quality function of each fragment map element, and calculating the credibility state of each registered map based on the credibility function;

[0009] According to the tentative state label and trusted state of each fragment map element, the final state label of each fragment map element is determined, and according to the final state label of each fragment map element, the reference mother map is updated based on each fragment map element.

[0010] According to a second aspect of the present invention, a high-precision map updating system is provided, comprising:

[0011] The registration module is used to register ground features based on the fragmented maps sampled multiple times and the reference master map to form a set of map feature registration pairs;

[0012] A first calculation module is configured to perform concentric circle pixel segmentation and clustering based on the center points of each element of the reference mother map, calculate the total distance loss between each pair of fragment map elements and the mother map element in the map element registration pair set, and tentatively assign a status label to each fragment map element based on the total distance loss, wherein the status label includes whether the newly added map element, the deleted map element, and the map element corresponding to the current mother map are of the same category;

[0013] Establish a module for establishing a corresponding quality function according to the confidence value of each fragment map element;

[0014] A second calculation module is used to calculate the credibility function of each state of each fragment map element based on the quality function of each fragment map element, and calculate the credibility state of each registered map based on the credibility function;

[0015] The update module is used to determine the final state label of each fragment map element according to the tentative state label and trusted state of each fragment map element, and update the reference mother map based on each fragment map element according to the final state label of each fragment map element.

[0016] According to a third aspect of the present invention, an electronic device is provided, comprising a memory and a processor, wherein the processor is configured to implement the steps of a high-precision map updating method when executing a computer management program stored in the memory.

[0017] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, on which a computer management program is stored. When the computer management program is executed by a processor, the steps of the high-precision map updating method are implemented.

[0018] This invention provides a high-precision map updating method and system. This method uses multiple sampling trajectories to register fixed control points for map fragments. It then performs segmentation and clustering based on the center points of each element in a reference master map. A quality function is established based on the confidence values ​​and spatial relationships of each map fragment's elements. Finally, the system integrates the quality functions and calculates the confidence scores to update map elements using the DS evidence theory. This method determines the positional relationship of elements within a map fragment with other samples based on their own position, providing important guidance for improving the accuracy and efficiency of geometric shape recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A flow chart of a high-precision map updating method provided by the present invention;

[0020] Figure 2 This is a schematic diagram of the overall process of the high-precision map update method;

[0021] Figure 3A schematic structural diagram of a high-precision map updating system provided by the present invention;

[0022] Figure 4 A schematic diagram of the hardware structure of a possible electronic device provided by the present invention;

[0023] Figure 5 A schematic diagram of the hardware structure of a possible computer-readable storage medium provided by the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, the technical features in the various embodiments or single embodiments provided by the present invention can be arbitrarily combined with each other to form a feasible technical solution. This combination is not restricted by the sequence of steps and / or structural composition mode, but must be based on the ability of ordinary technicians in this field to implement it. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0025] The focus is on solving the difficulties of how to classify and fuse a large number of map elements in map fragments after multiple sampling, and how to establish evaluation indicators. The present invention provides a flow chart of a high-precision map update method, such as Figure 1 Shown and Figure 2 As shown, the method includes:

[0026] S1, perform ground feature registration based on the fragmented maps sampled multiple times and the reference mother map to form a set of map feature registration pairs.

[0027] It is understandable that for the trajectory of multiple sampled map fragments, the fragment maps are aligned with the reference master map. Specifically, the Kabsch–Umeyama algorithm can be used to align the fragment maps with the reference master map, that is, to solve the least squares problem:

[0028]

[0029] Where y is the coordinate of the sub-image trajectory after the reference mother map is segmented, x is the coordinate of the sub-image trajectory after the registration fragment map is segmented, and (s, R, t) are the stretching parameter, rotation matrix and translation vector of the fragment map trajectory respectively.

[0030] Calculate the centroid of the trajectory points of the fragment map subgraph: where x i is the coordinate of the point of the fragment map subgraph trajectory, let It is the coordinate center of the point of the sub-graph trajectory of the parent map.

[0031] Find the covariance matrix And perform SVD decomposition C = ΣV T , find U and V T Then replace the middle ∑ with the diagonal matrix W, where all the diagonal elements except the last one are 1. If the last one is det(UV T )> is set to 1, otherwise it is -1.

[0032] Solve to get the optimal solution of R rotation matrix Then find the optimal solution of s in is the coordinate variance of the shape point of the fragment map subgraph trajectory, and finally the translation vector can be solved

[0033] According to (s, R, t), the fragment map elements are registered one by one with the reference mother map elements to obtain a map element registration pair set, that is, the map element registration pair set includes all paired fragment map elements and mother map elements.

[0034] S2, based on the center points of each element of the reference mother map, concentric circle pixel segmentation and clustering are performed, and the total distance loss of each pair of fragment map elements and the mother map elements in the map element registration pair set is calculated. Based on the total distance loss, the status label of each fragment map element is temporarily determined, and the status label includes newly added map elements, deleted map elements and map elements corresponding to the current mother map as the same category.

[0035] As an embodiment, the concentric circle pixel segmentation and clustering is performed based on the center points of each element of the reference mother map, and the total distance loss of each pair of fragment map elements and the mother map elements in the map element registration pair set is calculated, including: taking the centroid of each shape point string of the mother map element as the starting point, making a set of concentric circles with uniform radius gradually including all the shape points of the map element, and evenly dividing the concentric circles into multiple quadrants; traversing the mother map element shape point set, for each mother map element shape point, if the coordinates of the fragment map element shape point that is registered with it are within the same concentric circle radius and quadrant, then the distance between the two is calculated, and the fragment map elements that match the mother map element shape point are recorded. The number of primitive points; traverse each mother map element point in the mother map element point set, calculate the corresponding distance, count the number of matched points and the number of unmatched points, add up all distances to get the total distance of the current mother map element, add up the number of matched points to get the total number of pairs of the current mother map element, and add up the number of unmatched points to get the total number of unmatched points of the current mother map element; based on the total distance of the current mother map element, the total number of pairs of the current mother map element, and the total number of unmatched points of the current mother map element, calculate the total distance loss between the fragment map element and the mother map element based on the total distance loss function.

[0036] The traversal of the set of map element points of the mother map comprises: for each map element point of the mother map, if there is a fragment map element point whose coordinates are aligned with it and are within the same concentric circle radius and quadrant, then the distance between the two is calculated, and the number of fragment map element points that match the mother map element point is recorded, including:

[0037] For each mother map feature point in the mother map feature, if it is in the same concentric circle radius and quadrant, and there are multiple matching fragment map feature points, then calculate the distance between each matching fragment map feature point and the mother map feature point, and use the minimum distance as the distance d between the mother map feature point and the matching fragment map feature point. xy , and record the number n of fragment map feature points that match the main map feature points xy Cumulative increase: 1.

[0038] If the corresponding registered fragment map feature point cannot be found within the same concentric ring radius and quadrant, the distance d between the two xy Recorded as 0, the number of unpaired pairs is recorded as n y Cumulative increase: 1.

[0039] Correspondingly, the shape point distance d based on the matching of the current mother map element and the fragment map xy The cumulative distance d is summed up t, the total number of paired map elements in the current mother map, and the total number of unpaired map elements in the current mother map. The total distance loss between the fragment map elements and the mother map elements is calculated based on the total distance loss function, including:

[0040]

[0041] It can be understood that the aligned fragmented map elements and the map elements in the mother map are started from the centroid of each shape point string on the map element of the mother map, and concentric circles with a radius of r = 1.5 or 2 meters are drawn to gradually include all the shape points of the element, and then evenly divide it into 8 quadrants.

[0042] Traverse the set of map element points of the parent map. If the coordinates of the fragment map element points that are aligned with it are within the same concentric circle radius and quadrant, calculate the distance between them. The present invention can use but is not limited to Euclidean distance. The general calculation formula is:

[0043]

[0044] Of course, in addition to the traditional Euclidean distance, it can also be further mapped to a metric similar to the Mahalanobis distance:

[0045] d(x,y) ′ =d(x,y) T ∑ d d(x,y)tr(M) (3);

[0046] Where d(x,y) is the Euclidean distance, ∑ d is its covariance matrix, and tr() is the sum of the diagonal singular values ​​after SVD decomposition of the covariance matrix.

[0047] Formula (3) is used to calculate the distance between the feature points of the fragment map and the feature points of the mother map. The probability distribution is incorporated into the model from the perspective of Bayesian probability, taking into account the uncertainty in the observation process and using it as the proportion of the covariance matrix.

[0048] If the same registration map element has multiple points in the same ring and quadrant and the shape point of the parent map element, the minimum distance is used to define the current representative distance d xy , added to the total distance d of the current map element t , the current number of pairs is recorded as n xy , which is added to the total number of pairs of the current map elements n t On the contrary, if the corresponding registration map feature point cannot be found in the same ring and quadrant, the distance d xy Recorded as 0, the number of unpaired pairs is recorded as n yThe cumulative increase is 1. The total distance loss between each pair of registered fragment map elements and the parent map elements is calculated according to formula (1).

[0049] As an embodiment, based on the total distance loss, a status label of each fragment map element is temporarily determined. The status label includes newly added map elements, deleted map elements, and map elements corresponding to the current parent map, which are of the same category, including:

[0050] Determine the upper and lower boundaries of the distance threshold D Uppwr ,D lower , if the total distance loss D∈(D lower ,D Upper ), the fragment map element is temporarily regarded as the same type as the current parent map element and does not need to be updated. If D>D Upper , then the fragment map element is temporarily considered as a newly added map element. If D<D lower , then the fragmented map element is temporarily determined to be a map element that needs to be deleted.

[0051] It is understandable that the upper and lower boundaries of the threshold D are determined by the accumulated experience of mapping. Upper ,D lower , such as D∈(D lower ,D Upper ) then the map element is temporarily regarded as the same type as the map element corresponding to the current parent map. If it is above the upper limit, it is temporarily determined as a newly added map element; if it is below the lower limit, it is temporarily determined as a deleted map element.

[0052] S3, establishing a corresponding quality function according to the confidence value of each fragment map element.

[0053] As an embodiment, the method of establishing a corresponding quality function based on the confidence value of each fragment map element includes: establishing a probability distribution based on the geometric features and relative position relationship of each map element and calculating the confidence value of each map element point; and calculating the quality function of the fragment map element based on the confidence value of the registered fragment map element point:

[0054]

[0055] Where φ is an empty set, {a,,,} are all hidden states of credibility, conf represents the confidence value of the map feature point, and m(A) is the quality function of the fragment map feature.

[0056] Reconstructing the quality function m using the Deng entropy partition recognition framework * (A) is:

[0057]

[0058] Among them, |A| represents the number of hidden states of credibility.

[0059] It can be understood that a probability distribution is established based on the geometric features and relative positional relationships of each map element, and a confidence value of each map element point is calculated, where the confidence value ranges from [0, 1]. In practical applications, the present invention includes, but is not limited to, first mapping the quality function of the aligned map fragment points into five states. The expressions of the quality function of the fragment map elements are shown in Equations (4) and (5).

[0060] S4, based on the quality function of each fragment map element, calculating the credibility function of each state of each fragment map element, and based on the credibility function, calculating the credibility state of each registration map.

[0061] As an embodiment, the calculation of the credibility function of each state of each fragment map element based on the quality function of each fragment map element includes:

[0062] Based on the quality functions of the fragment map elements collected in two adjacent time periods in the order of collection time, the fusion quality function is calculated:

[0063]

[0064] in is the fusion operator, are the reconstruction quality functions of the fragment map features collected in two adjacent time periods.

[0065] Using the fused quality function Calculate the credibility function of each credibility hidden state:

[0066]

[0067] Among them, A j is a discrete state, i.e., a hidden state of credibility, and K is a discrete state of the fusion quality function A j The total number of I is the operator, when When it was established, Returns 1 when When it is not established, Return 0; based on the belief function of each discrete state, calculate the belief value of each discrete state.

[0068] As an embodiment, the method of determining the final state label of each fragment map element based on the tentative state label and trustworthy state of each fragment map element includes: dividing the discrete states of the fragment map element into trustworthy discrete states and untrustworthy discrete states; calculating the first trustworthiness value of the trustworthy discrete state and the second trustworthiness value of the untrustworthy discrete state based on the trustworthiness function; taking the trustworthiness of the discrete state corresponding to the larger of the first trustworthiness value and the second trustworthiness value as the trustworthy state of the fragment map element, and the trustworthy state represents the trustworthiness of the tentative state label of the fragment map element.

[0069] It can be understood that the four states of the above quality function are divided into credible discrete states and uncredible discrete states. For credible discrete states, their corresponding credibility values ​​are calculated based on the credibility function; for uncredible discrete states, their corresponding credibility values ​​are also calculated based on the credibility function. For example, let {a,b} be the state: if uncredible = {a,b}, then bel({a,b}) = m({a}) + m({b}) + m({a,b}) is its credibility value. If credible = {c,d}, then bel({c,d}) = m({c}) + m({d}) + m({c,d}) is its credibility value. Comparing the credibility values ​​of {a,b} and {c,d}, if the credibility value of {a,b} is greater than that of {c,d}, then the current fragment map element is tentatively labeled as uncredible; otherwise, the current fragment map element is tentatively labeled as credible.

[0070] S5, determining the final state label of each fragment map element according to the tentative state label and trusted state of each fragment map element, and updating the reference mother map based on each fragment map element according to the final state label of each fragment map element.

[0071] It is understood that the determination of whether to update a map element is based on the tentative status and the trustworthy status (trustworthy or untrustworthy) of the current fragmented map element. For example, if the current fragmented map element is tentatively designated as a new map element in step 2, but the corresponding trustworthy status is untrustworthy, the original map will remain unchanged.

[0072] See also Figure 3 , provides a high-precision map updating system, including a registration module 301, a first calculation module 302, a building module 303, a second calculation module 304 and an updating module 305, wherein:

[0073] A registration module 301 is used to perform ground feature registration based on the fragmented maps sampled multiple times and the reference mother map to form a set of map feature registration pairs;

[0074] The first calculation module 302 is configured to perform concentric circle pixel segmentation and clustering based on the center points of each element of the reference mother map, calculate the total distance loss between each pair of fragment map elements and the mother map element in the map element registration pair set, and tentatively assign a status label to each fragment map element based on the total distance loss. The status label includes whether the newly added map element, the deleted map element, and the map element corresponding to the current mother map are of the same class.

[0075] Establishing module 303, for establishing a corresponding quality function according to the confidence value of each fragment map element;

[0076] The second calculation module 304 is used to calculate the credibility function of each state of each fragment map element based on the quality function of each fragment map element, and calculate the credibility state of each registered map based on the credibility function;

[0077] The updating module 305 is used to determine the final state label of each fragment map element according to the tentative state label and trusted state of each fragment map element, and update the reference mother map based on each fragment map element according to the final state label of each fragment map element.

[0078] It can be understood that the high-precision map update system provided by the present invention corresponds to the high-precision map update method provided in the aforementioned embodiments. The relevant technical features of the high-precision map update system can refer to the relevant technical features of the high-precision map update method, which will not be repeated here.

[0079] See also Figure 4 , Figure 4 Schematic diagram of an embodiment of an electronic device provided by an embodiment of the present invention. Figure 4 As shown, an embodiment of the present invention provides an electronic device 400, including a memory 410, a processor 420, and a computer program 411 stored in the memory 410 and executable on the processor 420. When the processor 420 executes the computer program 411, the steps of the high-precision map updating method are implemented.

[0080] See also Figure 5 , Figure 5 Schematic diagram of an embodiment of a computer-readable storage medium provided by the present invention. Figure 5 As shown, this embodiment provides a computer-readable storage medium 500 on which a computer program 511 is stored. When the computer program 511 is executed by a processor, the steps of the high-precision map updating method are implemented.

[0081] Embodiments of the present invention provide a high-precision map updating method and system. This method uses multiple sampling trajectories to perform image registration of map fragments, fixing control points. It then performs segmentation and clustering based on the center points of each element in a reference master map. A quality function is established based on the confidence values ​​and spatial relationships of each map fragment's elements. Finally, the quality function is integrated using the DS evidence theory to calculate the confidence and update the map elements. This method addresses the challenges of classifying and fusing the large number of map elements in the multiple sampling map fragments and establishing evaluation metrics. By determining the positional relationship of the elements in the map fragments with other samples based on their own positions, this method provides important guidance for improving the accuracy and efficiency of geometric shape recognition.

[0082] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0083] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0084] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0085] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0086] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0087] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0088] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A high-precision map updating method, characterized in that: include: Perform ground feature registration based on the fragmented maps sampled multiple times and the reference master map to form a set of map feature registration pairs; Perform concentric circle pixel segmentation and clustering based on the center points of each element in the reference mother map, calculate the total distance loss between each pair of fragment map elements and the mother map elements in the map element registration pair set, and tentatively assign a status label to each fragment map element based on the total distance loss. The status label includes whether the newly added map element, the deleted map element, and the map element corresponding to the current mother map are of the same category. Establish a corresponding quality function based on the confidence value of each fragment map element; Calculating the credibility function of each state of each fragment map element based on the quality function of each fragment map element, and calculating the credibility state of each registered map based on the credibility function; Determine the final state label of each fragment map element based on the tentative state label and the trusted state of each fragment map element, and update the reference master map based on each fragment map element according to the final state label of each fragment map element; The step of establishing a corresponding quality function according to the confidence value of each fragment map element includes: The probability distribution is established based on the geometric characteristics and relative position relationship of each map element and the confidence value of each map element point is calculated; According to the confidence value of the registered fragment map feature points, the quality function of the fragment map feature is calculated: ； in is an empty set, All of them are hidden states of credibility. Represents the confidence value of the map feature point. is the quality function of the fragment map elements; Reconstructing the quality function using Deng entropy partitioning and identification framework for: ； in, The number of hidden states indicating the degree of credibility.

2. The high-precision map updating method according to claim 1, characterized in that: The process of performing trajectory registration on the fragmented map obtained by multiple samplings and the reference mother map to form a set of map element registration pairs includes: The fragment map is divided into several sub-maps using a sliding window or fixed cut method, and the Kabsch–Umeyama algorithm is used to align the fragment maps with the trajectory of the reference mother map, that is, to solve the least squares problem: ； in is the coordinate of the shape point of the trajectory in the reference mother graph, To align the coordinates of the shape points of the fragment map trajectory, They are the stretching parameters, rotation matrix and translation vector of the fragment map elements respectively; Calculate the centroid of the fragment map subgraph trajectory point string: ,in is the coordinate of the point of the fragment map subgraph trajectory, let The coordinate center of the shape point of the sub-graph trajectory of the parent map; Find the covariance matrix , and perform SVD decomposition on it , find and Then put the middle Substitute for a diagonal matrix , All the elements on the diagonal except the last one are 1, and if the last one If yes, it is set to 1, otherwise it is -1; Solved Optimal solution of the rotation matrix , then ask The optimal solution ,in is the coordinate variance of the shape point of the fragment map subgraph trajectory, and finally the translation vector can be solved ; according to The fragment map trajectories are registered with the reference master map trajectories one by one to obtain a set of map element registration pairs.

3. The high-precision map updating method according to claim 1, characterized in that: The concentric circle pixel segmentation and clustering based on the center points of each element of the reference mother map, and the calculation of the total distance loss between each pair of fragment map elements and the mother map element in the map element registration pair set, include: Taking the centroid of each shape point string of the map element of the mother map as the starting point, draw a set of concentric circles with evenly distributed radius to gradually include all the shape points of the map element, and evenly divide the concentric circles into multiple quadrants; Traverse the set of map feature points of the parent map. For each map feature point, if there is a fragment map feature point whose coordinates are aligned with it and are within the same concentric circle radius and quadrant, calculate the distance between them and record the number of fragment map feature points that match the map feature point. Traverse each mother map element shape point in the mother map element shape point set, calculate the corresponding distance, count the number of matched shape points and the number of unmatched shape points, accumulate all distances to obtain the total distance of the current mother map element, accumulate the number of matched shape points to obtain the total number of pairs of the current mother map element, and accumulate the number of unmatched shape points to obtain the total number of unmatched pairs of the current mother map element; Based on the total distance of the current mother map element, the total number of pairs of the current mother map element, and the total number of unpaired pairs of the current mother map element, the total distance loss between the fragment map element and the mother map element is calculated based on the total distance loss function.

4. The high-precision map updating method according to claim 3, characterized in that: The traversal of the set of map element points of the mother map comprises: for each map element point of the mother map, if the coordinates of a fragment map element point that matches the map element point are within the same concentric circle radius and quadrant, then the distance between the two is calculated, and the number of fragment map element points that match the map element point of the mother map is recorded, including: For each mother map feature point in the mother map feature, if it is in the same concentric circle radius and quadrant, and there are multiple matching fragment map feature points, then calculate the distance between each matching fragment map feature point and the mother map feature point, and use the minimum distance as the distance between the mother map feature point and the matching fragment map feature point. , and record the number of paired fragment map feature points that match the parent map feature points Cumulative increase 1; If the corresponding registered fragment map feature point cannot be found within the same concentric ring radius and quadrant, the distance between the two Recorded as 0, the number of unpaired pairs is recorded as Cumulative increase 1; Correspondingly, the shape point distance based on the matching of the current mother map element and the fragment map Cumulative distance formed by summation , the total number of paired map elements in the current mother map, and the total number of unpaired map elements in the current mother map. The total distance loss between the fragment map elements and the mother map elements is calculated based on the total distance loss function, including: 。 5. The high-precision map updating method according to claim 1 or 4, characterized in that: Based on the total distance loss, a status label is tentatively determined for each fragmented map element. The status label includes newly added map elements, deleted map elements, and map elements corresponding to the current parent map, which are of the same category, including: Determine the upper and lower boundaries of the distance threshold If the total distance loss between the parent map feature and the registered fragment map feature is , then the fragment map element is temporarily regarded as the same type as the current parent map element and does not need to be updated. , then the fragment map element is temporarily considered a newly added map element. If , then the fragmented map element is temporarily determined to be a map element that needs to be deleted.

6. The high-precision map updating method according to claim 1, characterized in that: The calculation of the credibility function of each state of each fragment map element based on the quality function of each fragment map element includes: Based on the quality functions of the fragment map elements collected in two adjacent time periods in the order of collection time, the fusion quality function is calculated: ； in is the fusion operator, are the reconstruction quality functions of the fragment map elements collected in two adjacent time periods; Using the fused quality function Calculate the credibility function of each credibility hidden state: ； in, is a discrete state, i.e., a hidden state of credibility, is the discrete state of the fusion quality function The total number of For operators, when hour, Returns 1 when hour, Return 0; Based on the belief function of each discrete state, the belief value of each discrete state is obtained.

7. The high-precision map updating method according to claim 6, characterized in that: The step of determining the final state label of each fragment map element based on the tentative state label and the trusted state of each fragment map element includes: Divide the discrete states of the fragment map elements into credible discrete states and uncredible discrete states; Based on the belief function, respectively calculating a first belief value of a credible discrete state and a second belief value of an uncredible discrete state; The credibility of the discrete state corresponding to the larger one of the first credibility value and the second credibility value is used as the credibility state of the fragment map element, and the credibility state represents the credibility of the temporary state label of the fragment map element.

8. A high-precision map updating system, characterized in that: include: The registration module is used to register ground features based on the fragmented maps sampled multiple times and the reference master map to form a set of map feature registration pairs; A first calculation module is configured to perform concentric circle pixel segmentation and clustering based on the center points of each element of the reference mother map, calculate the total distance loss between each pair of fragment map elements and the mother map element in the map element registration pair set, and tentatively assign a status label to each fragment map element based on the total distance loss, wherein the status label includes whether the newly added map element, the deleted map element, and the map element corresponding to the current mother map are of the same category; Establish a module for establishing a corresponding quality function according to the confidence value of each fragment map element; A second calculation module is used to calculate the credibility function of each state of each fragment map element based on the quality function of each fragment map element, and calculate the credibility state of each registered map based on the credibility function; An update module is used to determine the final state label of each fragment map element according to the tentative state label and the trusted state of each fragment map element, and update the reference mother map based on each fragment map element according to the final state label of each fragment map element; The step of establishing a corresponding quality function according to the confidence value of each fragment map element includes: The probability distribution is established based on the geometric characteristics and relative position relationship of each map element and the confidence value of each map element point is calculated; According to the confidence value of the registered fragment map feature points, the quality function of the fragment map feature is calculated: ； in is an empty set, All of them are hidden states of credibility. Represents the confidence value of the map feature point. is the quality function of the fragment map elements; Reconstructing the quality function using Deng entropy partitioning and identification framework for: ； in, The number of hidden states indicating the degree of credibility.

9. A computer-readable storage medium, characterized in that A computer management program is stored thereon, and when the computer management program is executed by the processor, the steps of the high-precision map updating method as described in any one of claims 1 to 6 are implemented.

Images