Quality evaluation method, device, medium and equipment for vehicle-side local map

By pre-processing and parameter calculation of local high-precision maps on the vehicle end, the map quality is comprehensively evaluated, and the one-sidedness and unity of the existing evaluation methods are solved, the credibility and availability of the map are improved, and the development of autonomous driving technology is promoted. .

CN118799837BActive Publication Date: 2024-11-15WUHAN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411273125.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-12
Publication Date
2024-11-15
Estimated Expiration
2044-09-12

AI Technical Summary

Technical Problem

The existing vehicle-end local high-precision map quality evaluation method is one-sided and single, and it is difficult to objectively and comprehensively describe the performance of map products, affecting the credibility and availability of map construction results.

Method used

By obtaining the prediction map and true map of the vehicle-end local map for preprocessing, completeness, logical consistency, location accuracy and thematic quality evaluation parameters are calculated, and quality evaluation results are generated, including Frecher distance, accuracy, and recall rate , balance F parameters and integrity F parameters, and comprehensively evaluate the performance of the map.

Benefits of technology

It realizes multi-angle and objective evaluation of local high-precision maps on the vehicle side, improves the credibility and usability of map products, guides the optimization of map construction methods, and promotes the application of autonomous driving technology.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118799837B_ABST
    Figure CN118799837B_ABST
Patent Text Reader

Abstract

The present application discloses a quality evaluation method, device, storage medium and electronic device for vehicle-side local maps. The method includes: obtaining a vehicle-side local map, the vehicle-side local map including a predicted map and a true map; preprocessing the vehicle-side local map; matching the preprocessed predicted map with the preprocessed true map to obtain a matching result; calculating the integrity evaluation parameters, logical consistency evaluation parameters, position accuracy evaluation parameters and thematic quality evaluation parameters of the vehicle-side local map based on the matching result; obtaining the quality evaluation result of the vehicle-side local map based on the integrity evaluation parameters, the logical consistency evaluation parameters, the position accuracy evaluation parameters and the thematic quality evaluation parameters. The present application can comprehensively and objectively describe the performance of map products.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence security technology, and in particular to a quality evaluation method, device, storage medium and electronic device for vehicle-side local maps. Background Art

[0002] As the market demand for automobile safety and intelligence continues to increase, the social and economic value of autonomous driving is becoming increasingly significant. The advancement and popularization of high-precision maps enable vehicles to achieve real-time and accurate positioning, while the widely used intelligent driving technology makes car driving simpler and smarter. The combination of high-precision maps and intelligent driving technology has jointly promoted the development of autonomous driving vehicle technology.

[0003] Traditional high-precision map construction solutions usually use mobile surveying vehicles to collect data and produce offline global high-precision maps based on simultaneous localization and mapping (SLAM). This solution has a complex process, high maintenance costs, and it is difficult to achieve full temporal and spatial coverage of roads.

[0004] Recently, the research on the construction of local high-precision maps on the vehicle side based on bird's-eye-view (BEV) perception has attracted widespread attention. It mainly relies on vehicle-mounted sensors such as cameras or lidars and deep learning technology, and uses surround images or laser point cloud data at the same time of acquisition to construct map elements such as lane dividers, road boundaries, and crosswalks within a certain range around the vehicle (usually 15m left and right, 30m front and back) in the vehicle coordinate system. The generated local high-precision map on the vehicle side contains semantic information such as road topology and traffic rules, providing a friendly vector map for downstream tasks of autonomous driving based on deep learning (such as motion prediction and planning), and is easy to expand.

[0005] However, due to differences in sensor performance and the influence of environmental factors such as weather, lighting, and occlusion during the mapping process, there are differences in the quality of local high-precision maps built on the vehicle side based on BEV perception. Therefore, it is crucial to conduct a reasonable quality evaluation of the mapping results, which helps to evaluate and compare the performance and robustness of the local high-precision map construction algorithms on the vehicle side and ensure the credibility and availability of the mapping data. In order to guide the optimization of mapping methods and promote their application in fields such as autonomous driving, it is of great significance to conduct quality evaluation of local high-precision maps on the vehicle side. However, the existing quality evaluation dimensions are one-sided and the methods are single, making it difficult to objectively and comprehensively describe the performance of map products. Summary of the invention

[0006] The embodiments of the present application provide a method, device, storage medium and electronic device for evaluating the quality of a vehicle-side local map, which can objectively and comprehensively describe the performance of a map product.

[0007] The embodiment of the present application provides a quality evaluation method for a vehicle-side local map, including:

[0008] Acquire a vehicle-side local map, wherein the vehicle-side local map includes a predicted map and a true map;

[0009] Preprocessing the vehicle-side local map;

[0010] Matching the preprocessed predicted map with the preprocessed true value map to obtain a matching result;

[0011] Calculating the integrity evaluation parameter, logic consistency evaluation parameter, location accuracy evaluation parameter and thematic quality evaluation parameter of the vehicle-side local map based on the matching result;

[0012] A quality evaluation result of the vehicle-side local map is obtained based on the integrity evaluation parameter, the logical consistency evaluation parameter, the position accuracy evaluation parameter and the thematic quality evaluation parameter.

[0013] Furthermore, in the above-mentioned vehicle-side local map quality evaluation method, the preprocessed predicted map is matched with the preprocessed true map to obtain a matching result, including:

[0014] Calculating the Fréchet distances between the set of predicted sequence elements in the predicted map and the set of true value sequence elements in the true value map one by one; wherein the set of predicted sequence elements includes a plurality of predicted map elements, and the set of true value sequence elements includes a plurality of true value map elements;

[0015] Determining the true map element having the smallest Fréchet distance to the set of predicted sequence elements;

[0016] If the minimum value of the Fréchet distance is less than a preset distance threshold, it is determined that the true value map element is the matching true value of the predicted map.

[0017] Furthermore, in the above-mentioned vehicle-side local map quality evaluation method, the integrity evaluation parameter of the vehicle-side local map is calculated based on the matching result, including:

[0018] Determine a first ratio of the unmatched predicted map elements to the total length of the predicted map elements based on the matching result, and use the first ratio as a redundant evaluation parameter;

[0019] Determine a second ratio of the total length of unmatched true map elements to the total length of true map elements based on the matching result, and use the second ratio as an omission evaluation parameter;

[0020] Determine precision and recall based on the predicted map and the true map;

[0021] An integrity evaluation parameter of the vehicle-side local map is determined based on the redundant evaluation parameter, the omitted evaluation parameter, the precision rate, and the recall rate.

[0022] Furthermore, in the above-mentioned vehicle-side local map quality evaluation method, the logical consistency evaluation parameter of the vehicle-side local map is calculated based on the matching result, including:

[0023] Sampling multiple random point pairs on the successfully matched true value map elements to obtain multiple pairs of sampling points;

[0024] If the sampling point can be projected onto the corresponding predicted map and the projection distance is less than a preset distance threshold, then the logical consistency evaluation parameter between the true map element and the predicted map element is determined to be 1.

[0025] Furthermore, in the above-mentioned vehicle-side local map quality evaluation method, the location accuracy evaluation parameter of the vehicle-side local map is calculated based on the matching result, including:

[0026] Calculating a first relative distance between the instance point in the predicted map element and the vehicle;

[0027] Calculating an absolute position error between an instance point in the predicted map element and a true map element matching the predicted map element;

[0028] Dividing the first relative distance into segments, and calculating an average value of the absolute position error in each segment;

[0029] calculating a second relative distance between the predicted map element and a first adjacent map element in a first adjacent map adjacent to the predicted map;

[0030] Calculating a third relative distance between the true value map element and a second adjacent map element in a second adjacent map adjacent to the true value map;

[0031] Obtaining a relative position error of the predicted map element based on the second relative distance and the third relative distance;

[0032] The position accuracy evaluation parameter of the vehicle-side local map is calculated based on the average value of the absolute position error and the relative position error.

[0033] Furthermore, in the above-mentioned vehicle-side local map quality evaluation method, the thematic quality evaluation parameter of the vehicle-side local map is calculated based on the matching result, including:

[0034] Among the predicted map elements that are matched successfully, calculating a third ratio of the number of the predicted map elements whose categories are consistent with the true map elements to the number of the predicted map elements that are matched successfully, and using the third ratio as a parameter for evaluating the correctness of thematic classification;

[0035] Among the successfully matched predicted map elements, calculating a fourth ratio of the number of predicted map elements whose directions are consistent with the directions of the true map elements to the number of successfully matched predicted map elements, and using the fourth ratio as a quantitative attribute accuracy evaluation parameter;

[0036] A topic quality evaluation parameter is obtained based on the topic classification correctness evaluation parameter and the quantitative attribute accuracy evaluation parameter.

[0037] Furthermore, in the above-mentioned quality evaluation method for the vehicle-side local map, the integrity parameter includes a balance F parameter and an integrity F parameter, and the integrity evaluation parameter of the vehicle-side local map is determined based on the redundant evaluation parameter, the omitted evaluation parameter, the precision rate and the recall rate, including:

[0038] The equilibrium F parameter and the integrity F parameter are calculated by the following formula:

[0039]

[0040]

[0041] in, is the accuracy, is the recall rate, is a redundant evaluation parameter, To omit the evaluation parameters, To balance the F parameter, is the integrity F parameter.

[0042] The embodiment of the present application also provides a quality evaluation device for a vehicle-side local map, including:

[0043] An acquisition module, used to acquire a vehicle-side local map, wherein the vehicle-side local map includes a predicted map and a true map;

[0044] A preprocessing module, used for preprocessing the vehicle-side local map;

[0045] A matching module, used for matching the preprocessed predicted map with the preprocessed true value map to obtain a matching result;

[0046] A calculation module, used for calculating the integrity evaluation parameter, the logic consistency evaluation parameter, the position accuracy evaluation parameter and the thematic quality evaluation parameter of the vehicle-side local map based on the matching result;

[0047] The quality evaluation module is used to obtain the quality evaluation result of the vehicle-side local map based on the integrity evaluation parameter, the logical consistency evaluation parameter, the position accuracy evaluation parameter and the thematic quality evaluation parameter.

[0048] An embodiment of the present application also provides a computer-readable storage medium, in which a plurality of instructions are stored, and the instructions are suitable for being loaded by a processor to execute any of the above-mentioned quality assessment methods for vehicle-side local maps.

[0049] An embodiment of the present application also provides an electronic device, including a processor and a memory, wherein the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used for the steps in the quality assessment method for a vehicle-side local map as described in any one of the above items.

[0050] The quality evaluation method, device, storage medium and electronic device for vehicle-side local maps provided in this application evaluate the quality of vehicle-side local maps by calculating integrity evaluation parameters, logical consistency evaluation parameters, position accuracy evaluation parameters and thematic quality evaluation parameters. The quality of vehicle-side local maps is evaluated from multiple angles, and the performance of map products can be comprehensively and objectively described. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The technical solution and other beneficial effects of the present application will be made apparent by describing in detail the specific implementation methods of the present application in conjunction with the accompanying drawings.

[0052] Figure 1 A flowchart of a method for evaluating the quality of a vehicle-side local map provided in an embodiment of the present application.

[0053] Figure 2 Another flowchart of the quality assessment method for vehicle-side local maps provided in an embodiment of the present application.

[0054] Figure 3 A schematic diagram of topological consistency determination provided in an embodiment of the present application.

[0055] Figure 4 This is a distribution curve diagram of the change in the absolute position error of the predicted map element relative to the vehicle distance provided in the embodiment of the present application.

[0056] Figure 5 A schematic diagram of a basis for determining the accuracy of a directional attribute provided in an embodiment of the present application.

[0057] Figure 6 A schematic diagram of the structure of a quality assessment device for a vehicle-side local map provided in an embodiment of the present application.

[0058] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0059] Figure 8 Another schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0061] The embodiment of the present application provides a method, device, storage medium and electronic device for evaluating the quality of a vehicle-side local map. The embodiment of the present application provides a device for evaluating the quality of a vehicle-side local map, which can be integrated into an electronic device, which can be a terminal, a server or other device, wherein the terminal can include a tablet computer, a laptop computer, a personal computer (PC), a micro processing box, or other devices.

[0062] See also Figure 1 and Figure 2 , Figure 1 A flowchart of a method for evaluating the quality of a vehicle-side local map provided in an embodiment of the present application. Figure 2 Another flow chart of the quality evaluation method for a vehicle-side local map provided in an embodiment of the present application, which is applied to an electronic device, the quality evaluation method for a vehicle-side local map includes the following steps:

[0063] S1, obtain the local map of the vehicle side, which includes the predicted map and the true map.

[0064] Specifically, obtain a Pickle file (Pickle is a persistence module dedicated to Python. A pickle file is a binary file that serializes data and its structure. It is a commonly used data storage format in Python). The Pickle file includes a set of predicted map feature instance points and a set of map feature instance points of the true map.

[0065] S2, preprocessing the local map on the vehicle side.

[0066] Specifically, the Pickle file is converted into a JSON (JavaScript Object Notation, a lightweight data exchange format) file for subsequent quality evaluation. Specifically, first use the pickle module to read the data in the Pickle file, and then convert the NumPy (Numerical Python, an open source numerical computing extension of Python) array into a Python built-in type. Next, traverse each frame of the data, extract the key information, organize the map feature instances of the predicted map and the true value map into a dictionary form in a specific format, and store them in new dictionaries respectively. Finally, use the json module to write the dictionary into the JSON files of the predicted map and the true value map.

[0067] S3, matching the preprocessed prediction map with the preprocessed true value map to obtain a matching result.

[0068] In one embodiment, step S3 includes the following steps:

[0069] S31, calculating the Fréchet distance between the set of predicted sequence elements in the predicted map and the set of true value sequence elements in the true value map one by one; wherein the set of predicted sequence elements includes a number of predicted map elements, and the set of true value sequence elements includes a number of true value map elements.

[0070] Specifically, the discrete version of the Fréchet distance is used to evaluate the predicted sequence feature set and the set of true value sequence elements The geometric similarity between . express a sequence of line segment endpoints, From the original input The uniform sampling has A sequence of vertices, where and between Each position of can be approximated by using an affine transformation, namely .

[0071] set up and is a broken line, and is the corresponding element set sequence. yes and Sequence of different pairs between:

[0072]

[0073] index and is a non-decreasing surjection, satisfying , , , , and for all , both and .

[0074] definition The norm of for The length of the longest pair in

[0075]

[0076] and The discrete Fréchet distance between is defined as:

[0077]

[0078] in, yes and The coupling between them.

[0079] S32, determining the true map element having the smallest Fréchet distance with the predicted sequence element set.

[0080] S33: If the minimum value of the Fréchet distance is less than a preset distance threshold, it is determined that the true value map element is a matching true value of the predicted map.

[0081] Specifically, if the Fréchet distance value is lower than a preset distance threshold (generally 0.5, 1.0 or 1.5), then the map truth element is considered to be the matching truth value of the current predicted map. If the Fréchet distance value is greater than or equal to the preset distance threshold, then the current predicted map has no matching truth map element.

[0082] S4, based on the matching results, calculate the integrity evaluation parameters, logical consistency evaluation parameters, location accuracy evaluation parameters and thematic quality evaluation parameters of the vehicle-side local map.

[0083] It should be noted that the local high-precision map on the vehicle side uses on-board sensors (such as cameras, lidar, etc.) to collect data within a limited range around the vehicle at a certain moment, encodes image or point cloud features with the help of deep learning and other technologies, and then directly outputs map elements for real-time decision-making and regulatory control needs, with the characteristics of real-time and locality.

[0084] The conventional data structure contains the geometric position of the map element, the number of points that constitute the instance, and the category information. Specifically, sample_token is the unique record identifier of the current frame, and the different types of data collected by various sensors at a certain timestamp are assigned the same sample_token; vectors is the vector point set of the map element of the current frame, including the geometric coordinates of each map element in the vehicle coordinate system (pts), the number of points required to form the map element instance (pts_num); the category name of the map element (cls_name) and the category identifier (type). Depending on the data processing method, the above data structure may show certain changes, but the information contained therein is roughly the same.

[0085] According to the above characteristics, the quality element group of the local high-precision map on the vehicle side is defined as four categories: completeness, logical consistency, position accuracy, and thematic quality. The corresponding quality elements retain the redundancy and omission of completeness, the topological consistency of logical consistency, the absolute position accuracy and relative position accuracy of position accuracy, and the thematic classification correctness and quantitative attribute accuracy of thematic quality.

[0086] In one embodiment, calculating the integrity evaluation parameter of the vehicle-side local map based on the matching result includes the following steps:

[0087] S411, determining a first ratio of the unmatched predicted map elements to the total length of the predicted map elements based on the matching result, and using the first ratio as a redundant evaluation parameter.

[0088] It can be specifically expressed by the following formula:

[0089]

[0090] in, represents redundant evaluation parameters, Represents the total length of unmatched predicted map features. Represents the total length of the matched predicted map elements. To predict the total length of map features.

[0091] S412: Determine a second ratio of the total length of unmatched true map elements to the total length of true map elements based on the matching result, and use the second ratio as a missing evaluation parameter.

[0092] It can be specifically expressed by the following formula:

[0093]

[0094] in, represents the missing evaluation parameter, Represents the total length of unmatched ground truth map features, Represents the total length of the matched true map elements, It is the total length of the true value map feature.

[0095] S413, determining the precision and recall according to the predicted map and the true value map.

[0096] Specifically, precision is the ratio of predicted maps that are actually positive examples to those that are actually positive examples. Recall is the ratio of predicted maps that are actually positive examples to those that are predicted to be positive examples.

[0097] S414, determining integrity evaluation parameters of the vehicle-side local map based on redundant evaluation parameters, omitted evaluation parameters, precision and recall.

[0098] Specifically, the integrity parameter includes a balance F parameter and an integrity F parameter. Step S414 includes:

[0099] The equilibrium F parameter and the integrity F parameter are calculated by the following formula:

[0100]

[0101]

[0102] in, is the accuracy, is the recall rate, is a redundant evaluation parameter, To omit the evaluation parameters, To balance the F parameter, is the integrity F parameter.

[0103] In one embodiment, calculating the logic consistency evaluation parameter of the vehicle-side local map based on the matching result includes the following steps:

[0104] S421, sampling multiple random point pairs on the successfully matched true value map elements to obtain multiple pairs of sampling points.

[0105] S422: If the sampling point can be projected onto the corresponding predicted map and the projection distance is less than a preset distance threshold, the logical consistency evaluation parameter between the true map element and the predicted map element is determined to be 1, otherwise it is 0.

[0106] Specifically, in order to evaluate the logical consistency of the local high-precision map on the vehicle side, the connectivity of map elements is selected as the measure of topological consistency (also known as logical consistency).

[0107] This evaluation index measures the degree of consistency between the connectivity of the predicted map elements and the true value. Specifically, repeated random point pairs are sampled on the successfully matched true value map elements. If the sampling points can be projected onto the corresponding predicted map elements and the projection distance is within a certain range, the topology is considered to be connected and the connectivity of the predicted map elements is consistent with the true value; otherwise, the connectivity of the predicted map elements is considered inconsistent with the true value. Figure 3 A schematic diagram of topology consistency determination provided in an embodiment of the present application is shown in FIG. Figure 3 As shown, the topological consistency is measured by calculating the proportion of predicted map features that are consistent with the true map features in connectivity. The calculation formula is as follows:

[0108]

[0109] in, represents the topological consistency evaluation parameter (that is, the logical consistency evaluation parameter), Indicates the number of connectivity values ​​consistent with the true value in the predicted map features that are matched successfully. Indicates the number of predicted map features that were matched successfully.

[0110] In one embodiment, calculating the location accuracy evaluation parameter of the vehicle-side local map based on the matching result includes the following steps:

[0111] S431, calculating the first relative distance between the instance point in the predicted map element and the vehicle (i.e., the center of the local high-precision map on the vehicle side).

[0112] Considering that map features are usually long and narrow, it may not be practical to directly measure the distance of the predicted map feature as a whole relative to the vehicle. Therefore, we choose to calculate the distance of each instance point in the predicted map feature relative to the vehicle and calculate the absolute position error between the instance point and its matching true value feature.

[0113] Assume that the predicted map element is A collection of points , and the matching true map element is A collection of points , then predict the instance points in the map features The distance relative to the vehicle is as follows:

[0114]

[0115] in, Represents the instance point in the predicted map feature The first relative distance to the vehicle, Represents predicted map features Middle The coordinates of a point.

[0116] S432, calculating the absolute position error between the instance point in the predicted map element and the true value map element matching the predicted map element.

[0117] Specifically, the example point The absolute position error between the map element and its matching ground truth can be expressed as follows:

[0118]

[0119] in, Represents predicted map features Middle The coordinates of the points, Represents the true map feature that matches the predicted map feature Middle The coordinates of the points, That is the point To match the true map feature The absolute position error.

[0120] S433, dividing the first relative distance into segments, and calculating an average value of the absolute position error in each segment.

[0121] It should be noted that in order to more accurately evaluate the distribution of the absolute position error of the predicted map element relative to the change in the distance from the vehicle, the distance of the predicted map element instance point relative to the vehicle is divided into segments, and the average value of the absolute position error between all predicted map element instance points and their matching true value elements is calculated in each segment.

[0122] S434, calculating a second relative distance between the predicted map element and a first adjacent map element in a first adjacent map adjacent to the predicted map.

[0123] S435: Calculate a third relative distance between the true value map element and a second adjacent map element in a second adjacent map adjacent to the true value map.

[0124] S436: Obtain a relative position error of the predicted map element based on the second relative distance and the third relative distance.

[0125] Figure 4 This is a distribution curve diagram of the change in the absolute position error of the predicted map element relative to the vehicle distance provided in the embodiment of the present application. Specifically, the relative distance between the lane separation lines is mainly calculated, and the specific calculation formula is as follows:

[0126]

[0127]

[0128]

[0129] in, Represents predicted map features Its adjacent map features The relative distance between Represents true map features Its adjacent map features The relative distance between and Represents the predicted map elements and ground truth map elements The coordinates of and Respectively and The coordinates of the adjacent map features; Indicates the number of map features. Represents the relative position error of the predicted map feature.

[0130] S437, calculating the position accuracy evaluation parameter of the vehicle-side local map based on the average value of the absolute position error and the relative position error.

[0131] The existing vehicle-side local high-precision map quality evaluation method already includes the evaluation content of the correctness of thematic classification in thematic quality evaluation. This evaluation method first calculates the AP of each category, and then averages the AP of each category to obtain mAP. This evaluation index calculates the AP and mAP of each category to measure the degree of match between the predicted map elements and the true map elements. It can take into account the correctness of thematic classification of map elements in the map, but the map elements in the vehicle-side local high-precision map also imply attribute information such as direction, and it is necessary to combine other evaluation indicators to comprehensively evaluate the thematic quality of the map.

[0132] In one embodiment, calculating the thematic quality evaluation parameter of the vehicle-side local map based on the matching result includes the following steps:

[0133] S441, among the successfully matched predicted map elements, calculate the third ratio of the number of predicted map elements whose categories are consistent with the categories of the true map elements to the number of successfully matched predicted map elements, and use the third ratio as a parameter for evaluating the correctness of thematic classification.

[0134] Specifically, the calculation formula is as shown below. The closer the ratio is to 1, the more the number of predicted map element categories that are consistent with the true values ​​is, indicating that the predicted category attributes are more accurate.

[0135]

[0136] in, represents the evaluation parameter of the correctness of thematic classification, Indicates the number of correctly predicted category attributes in the successfully matched predicted map features. Indicates the number of predicted map features that were matched successfully.

[0137] S442, among the successfully matched predicted map elements, calculate a fourth ratio of the number of predicted map elements whose directions are consistent with the directions of the true map elements to the number of successfully matched predicted map elements, and use the fourth ratio as a quantitative attribute accuracy evaluation parameter.

[0138] Figure 5 A schematic diagram of the direction attribute accuracy judgment basis provided in the embodiment of the present application is shown in FIG. Figure 5 As shown in the figure, first, the average direction of the predicted map elements and the true map elements that are matched successfully are calculated respectively, and then the difference between the average direction of the two is compared with the specified threshold. If the difference between the two is less than or equal to the preset threshold, it is considered that the direction of the predicted map element is consistent with that of the true map element; otherwise, it is considered that the directions are inconsistent. Finally, the proportion of the predicted map elements whose directions are consistent with the true values ​​is calculated among the successfully matched predicted map elements. The calculation formula is shown in the following formula.

[0139]

[0140] in, represents the direction attribute accuracy evaluation parameter, Indicates the number of directions of the predicted map features that are matched successfully and are consistent with the true values. Indicates the number of predicted map features that were matched successfully.

[0141] S443, obtaining a topic quality evaluation parameter based on the topic classification correctness evaluation parameter and the quantitative attribute accuracy evaluation parameter.

[0142] S5, obtaining a quality evaluation result of the vehicle-side local map based on the integrity evaluation parameters, the logical consistency evaluation parameters, the position accuracy evaluation parameters and thematic quality evaluation parameters.

[0143] This application evaluates the quality of the vehicle-side local map by calculating integrity evaluation parameters, logical consistency evaluation parameters, location accuracy evaluation parameters and thematic quality evaluation parameters. It evaluates the quality of the vehicle-side local map from multiple angles and can comprehensively and objectively describe the performance of the map product.

[0144] According to the method described in the above embodiment, this embodiment will be further described from the perspective of a quality evaluation device for a vehicle-side local map. The quality evaluation device for a vehicle-side local map can be implemented as an independent entity or integrated in an electronic device. The electronic device can be a terminal, a server, or other equipment. The terminal can include a tablet computer, a laptop computer, a personal computer (PC), a micro processing box, or other equipment.

[0145] See also Figure 6 , Figure 6 The quality evaluation device for a vehicle-side local map provided by an embodiment of the present application is specifically described, and is applied to an electronic device. The quality evaluation device for a vehicle-side local map may include:

[0146] An acquisition module is used to acquire a local map on the vehicle side, where the local map on the vehicle side includes a predicted map and a true map;

[0147] A preprocessing module is used to preprocess the local map on the vehicle side;

[0148] A matching module is used to match the preprocessed prediction map with the preprocessed true value map to obtain a matching result;

[0149] A calculation module, used to calculate the integrity evaluation parameters, logical consistency evaluation parameters, location accuracy evaluation parameters and thematic quality evaluation parameters of the vehicle-side local map based on the matching results;

[0150] The quality evaluation module is used to obtain the quality evaluation result of the vehicle-side local map based on integrity evaluation parameters, logical consistency evaluation parameters, position accuracy evaluation parameters and thematic quality evaluation parameters.

[0151] During specific implementation, the above modules and / or units can be implemented as independent entities, or can be arbitrarily combined to be implemented as the same or several entities. The specific implementation of the above modules and / or units can refer to the previous method embodiments. The specific beneficial effects that can be achieved can also refer to the beneficial effects in the previous method embodiments, which will not be repeated here.

[0152] In addition, an embodiment of the present application also provides an electronic device, which may be a computer, a tablet computer, or other device. Figure 7 As shown, the electronic device 400 includes a processor 401 and a memory 402. The processor 401 is electrically connected to the memory 402.

[0153] The processor 401 is the control center of the electronic device 400. It uses various interfaces and lines to connect various parts of the entire electronic device. It executes various functions of the electronic device and processes data by running or loading applications stored in the memory 402 and calling data stored in the memory 402, thereby monitoring the electronic device as a whole.

[0154] In this embodiment, the processor 401 in the electronic device 400 loads instructions corresponding to the processes of one or more application programs into the memory 402 according to the following steps, and the processor 401 runs the application programs stored in the memory 402 to implement various functions:

[0155] Acquire a vehicle-side local map, wherein the vehicle-side local map includes a predicted map and a true map;

[0156] Preprocessing the vehicle-side local map;

[0157] Matching the preprocessed predicted map with the preprocessed true value map to obtain a matching result;

[0158] Calculating the integrity evaluation parameter, logic consistency evaluation parameter, location accuracy evaluation parameter and thematic quality evaluation parameter of the vehicle-side local map based on the matching result;

[0159] A quality evaluation result of the vehicle-side local map is obtained based on the integrity evaluation parameter, the logical consistency evaluation parameter, the position accuracy evaluation parameter and the thematic quality evaluation parameter.

[0160] The electronic device can implement the steps in any embodiment of the quality evaluation method for the vehicle-side local map provided in the embodiments of the present application. Therefore, it can achieve the beneficial effects that can be achieved by any quality evaluation method for the vehicle-side local map provided in the embodiments of the present invention. Please refer to the previous embodiments for details and will not be repeated here.

[0161] Figure 8 The specific structural block diagram of the electronic device provided in the embodiment of the present invention is shown, and the electronic device can be used to implement the quality evaluation method for the vehicle-side local map provided in the above embodiment. The electronic device 500 can be a terminal, a server, etc., wherein the terminal can include a tablet computer, a laptop computer, a personal computer (PC, Personal Computer), a micro processing box, or other devices.

[0162] The RF circuit 510 is used to receive and send electromagnetic waves, realize the mutual conversion between electromagnetic waves and electrical signals, and thus communicate with a communication network or other devices. The RF circuit 510 may include various existing circuit elements for performing these functions, such as antennas, radio frequency transceivers, digital signal processors, encryption / decryption chips, user identity module (SIM) cards, memories, etc. The RF circuit 510 can communicate with various networks such as the Internet, corporate intranets, wireless networks, or communicate with other devices through wireless networks. The above-mentioned wireless networks may include cellular telephone networks, wireless local area networks, or metropolitan area networks. The above-mentioned wireless networks may use various communication standards, protocols and technologies, including but not limited to Global System for Mobile Communication (GSM), Enhanced Data GSM Environment (EDGE), Wideband Code Division Multiple Access (WCDMA), Code Division Access (CDMA), Time Division Multiple Access (TDMA), Wireless Fidelity (Wi-Fi) (such as Institute of Electrical and Electronics Engineers standards IEEE 802.11a, IEEE 802.11b, IEEE802.11g and / or IEEE802.11n), Voice over Internet Protocol (VoIP), Worldwide Interoperability for Microwave Access (Wi-Max), other protocols for email, instant messaging and short messages, and any other suitable communication protocols, even those that have not yet been developed.

[0163] The memory 520 can be used to store software programs and modules, such as the corresponding program instructions / modules in the above embodiments. The processor 580 executes various functional applications and data processing by running the software programs and modules stored in the memory 520, that is, realizing the functions of taking pictures with the front camera, processing the captured images, and switching the display color of the display content on the display screen. The memory 520 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 520 may further include a memory remotely arranged relative to the processor 580, and these remote memories may be connected to the electronic device 500 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0164] The input unit 530 may be used to receive input digital or character information, and generate a keyboard and a mouse related to user settings and function control.

[0165] The display unit 540 may be used to display information input by the user or information provided to the user and various graphical user interfaces, which may be composed of graphics, text, icons, videos, and any combination thereof. The display unit 540 may include a display panel 541, which may be configured in the form of an LCD (Liquid Crystal Display), an OLED (Organic Light-Emitting Diode), or the like.

[0166] The audio circuit 560, the speaker 561, and the microphone 562 can provide an audio interface between the user and the electronic device 500. The audio circuit 560 can transmit the electrical signal converted from the received audio data to the speaker 561, which is converted into a sound signal for output; on the other hand, the microphone 562 converts the collected sound signal into an electrical signal, which is received by the audio circuit 560 and converted into audio data, and then the audio data is output to the processor 580 for processing, and then sent to another terminal through the RF circuit 510, or the audio data is output to the memory 520 for further processing. The audio circuit 560 may also include an earplug jack to provide communication between an external headset and the electronic device 500.

[0167] The electronic device 500 can help the user receive requests, send information, etc. through the transmission module 570 (such as a Wi-Fi module), which provides the user with wireless broadband Internet access. Although the transmission module 570 is shown in the figure, it can be understood that it is not a necessary component of the electronic device 500 and can be omitted as needed without changing the essence of the invention.

[0168] The processor 580 is the control center of the electronic device 500. It uses various interfaces and lines to connect various parts of the entire mobile phone. By running or executing software programs and / or modules stored in the memory 520, and calling data stored in the memory 520, it executes various functions of the electronic device 500 and processes data, thereby monitoring the electronic device as a whole. Optionally, the processor 580 may include one or more processing cores; in some embodiments, the processor 580 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, etc., and the modem processor mainly processes wireless communications. It can be understood that the above-mentioned modem processor may not be integrated into the processor 580.

[0169] The electronic device 500 also includes a power supply 590 (such as a battery) for supplying power to various components. In some embodiments, the power supply can be logically connected to the processor 580 through a power management system, so that the power management system can manage charging, discharging, and power consumption management. The power supply 590 can also include any components such as one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, and power status indicators.

[0170] Although not shown, the electronic device 500 also includes a camera (such as a front camera, a rear camera), a Bluetooth module, etc., which will not be described in detail here. Specifically in this embodiment, the display unit of the electronic device is a touch screen display, and the mobile terminal also includes a memory, and one or more programs, wherein one or more programs are stored in the memory, and are configured to be executed by one or more processors. One or more programs include instructions for performing the following operations:

[0171] Acquire a vehicle-side local map, wherein the vehicle-side local map includes a predicted map and a true map;

[0172] Preprocessing the vehicle-side local map;

[0173] Matching the preprocessed predicted map with the preprocessed true value map to obtain a matching result;

[0174] Calculating the integrity evaluation parameter, logic consistency evaluation parameter, location accuracy evaluation parameter and thematic quality evaluation parameter of the vehicle-side local map based on the matching result;

[0175] A quality evaluation result of the vehicle-side local map is obtained based on the integrity evaluation parameter, the logical consistency evaluation parameter, the position accuracy evaluation parameter and the thematic quality evaluation parameter.

[0176] In specific implementation, the above modules can be implemented as independent entities, or can be arbitrarily combined and implemented as the same or several entities. The specific implementation of the above modules can be found in the previous method embodiments, which will not be repeated here.

[0177] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructions, or by controlling related hardware through instructions, and the instructions can be stored in a computer-readable storage medium and loaded and executed by a processor. To this end, an embodiment of the present invention provides a storage medium, which stores multiple instructions, and the instructions can be loaded by a processor to execute the steps of any embodiment of the quality assessment method for a vehicle-side local map provided by an embodiment of the present invention.

[0178] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0179] Since the instructions stored in the storage medium can execute the steps in any embodiment of the quality evaluation method for the vehicle-side local map provided in the embodiments of the present invention, the beneficial effects that can be achieved by any quality evaluation method for the vehicle-side local map provided in the embodiments of the present invention can be achieved. Please refer to the previous embodiments for details and will not be repeated here.

[0180] The above is a detailed introduction to a quality evaluation method, device, storage medium and electronic device for a vehicle-side local map provided in an embodiment of the present application. This article uses specific examples to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A quality evaluation method for a vehicle-side local map, characterized in that: The method comprises: Acquire a vehicle-side local map, wherein the vehicle-side local map includes a predicted map and a true map; Preprocessing the vehicle-side local map; Matching the preprocessed prediction map with the preprocessed true value map to obtain a matching result; wherein, matching the preprocessed prediction map with the preprocessed true value map to obtain a matching result includes: calculating the Fréchet distance between the prediction sequence element set in the prediction map and the true value sequence element set in the true value map one by one; wherein, the prediction sequence element set includes a plurality of prediction map elements, and the true value sequence element set includes a plurality of true value map elements; determining the true value map element having the smallest Fréchet distance with the prediction sequence element set; if the minimum Fréchet distance is less than a preset distance threshold, determining that the true value map element is the matching true value of the prediction map; Based on the matching result, the integrity evaluation parameter, logical consistency evaluation parameter, location accuracy evaluation parameter and thematic quality evaluation parameter of the vehicle-side local map are calculated; wherein, the calculation of the integrity evaluation parameter of the vehicle-side local map based on the matching result includes: determining a first ratio of unmatched predicted map elements to the total length of predicted map elements based on the matching result, and using the first ratio as a redundant evaluation parameter; determining a second ratio of the total length of unmatched true map elements to the total length of true map elements based on the matching result, and using the second ratio as an omission evaluation parameter; determining a precision rate and a recall rate based on the predicted map and the true map; determining the integrity evaluation parameter of the vehicle-side local map based on the redundant evaluation parameter, the omission evaluation parameter, the precision rate and the recall rate; The integrity parameter includes a balance F parameter and an integrity F parameter, and the determining of the integrity evaluation parameter of the vehicle-side local map based on the redundant evaluation parameter, the omitted evaluation parameter, the precision rate and the recall rate includes: The equilibrium F parameter and the integrity F parameter are calculated by the following formula: in, is the accuracy, is the recall rate, is a redundant evaluation parameter, To omit the evaluation parameters, To balance the F parameter, is the integrity F parameter; Calculating the logic consistency evaluation parameter of the vehicle-side local map based on the matching result, including: performing multiple random point pair sampling on the successfully matched true value map element to obtain multiple pairs of sampling points; if the sampling point can be projected onto the corresponding predicted map, and the projection distance is less than a preset distance threshold, determining that the logic consistency evaluation parameter of the true value map element and the predicted map element is 1; Calculating the position accuracy evaluation parameter of the vehicle-side local map based on the matching result, including: calculating the first relative distance between the instance point in the predicted map element and the vehicle; calculating the absolute position error between the instance point in the predicted map element and the true map element matching the predicted map element; segmenting the first relative distance, and calculating the average value of the absolute position error in each segment; calculating the second relative distance between the predicted map element and a first adjacent map element in a first adjacent map adjacent to the predicted map; calculating the third relative distance between the true map element and a second adjacent map element in a second adjacent map adjacent to the true map; obtaining the relative position error of the predicted map element based on the second relative distance and the third relative distance; calculating the position accuracy evaluation parameter of the vehicle-side local map based on the average value of the absolute position error and the relative position error; Calculating the thematic quality evaluation parameters of the vehicle-side local map based on the matching results, including: among the successfully matched predicted map elements, calculating a third ratio of the number of predicted map elements whose categories are consistent with the categories of the true map elements to the number of successfully matched predicted map elements, and using the third ratio as a thematic classification correctness evaluation parameter; among the successfully matched predicted map elements, calculating a fourth ratio of the number of predicted map elements whose directions are consistent with the directions of the true map elements to the number of successfully matched predicted map elements, and using the fourth ratio as a quantitative attribute accuracy evaluation parameter; obtaining the thematic quality evaluation parameters based on the thematic classification correctness evaluation parameters and the quantitative attribute accuracy evaluation parameters; A quality evaluation result of the vehicle-side local map is obtained based on the integrity evaluation parameter, the logical consistency evaluation parameter, the position accuracy evaluation parameter and the thematic quality evaluation parameter.

2. A quality assessment device for a vehicle-side local map, the quality assessment device for a vehicle-side local map is used to implement the quality assessment method for a vehicle-side local map according to claim 1, characterized in that: include: An acquisition module, used to acquire a vehicle-side local map, wherein the vehicle-side local map includes a predicted map and a true map; A preprocessing module, used for preprocessing the vehicle-side local map; A matching module, used for matching the preprocessed predicted map with the preprocessed true value map to obtain a matching result; A calculation module, used for calculating the integrity evaluation parameter, the logic consistency evaluation parameter, the position accuracy evaluation parameter and the thematic quality evaluation parameter of the vehicle-side local map based on the matching result; The quality evaluation module is used to obtain the quality evaluation result of the vehicle-side local map based on the integrity evaluation parameter, the logical consistency evaluation parameter, the position accuracy evaluation parameter and the thematic quality evaluation parameter.

3. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the quality assessment method for vehicle-side local maps as described in claim 1.

4. An electronic device, characterized in that: It includes a processor and a memory, the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used to execute the steps in the quality assessment method for vehicle-side local maps described in claim 1.