Map quality test method, computer-readable storage medium, and intelligent device

Abstract

The present application relates to the technical field of automated driving, and particularly, provided are a map quality test method, a computer-readable storage medium and an intelligent device, which aim to solve the problems that existing high-definition maps possibly have errors and thus affect safety of driving. For this purpose, the map quality test method of the present application comprises: on the basis of map data of a target area, acquiring first attribute information of a map element to be tested; on the basis of sensed data of the target area, acquiring second attribute information of a reference map element; associating the reference map element with the map element to be tested; comparing the first attribute information with the second attribute information; and at least according to the comparison result, determining a map quality test result of the target area. The method performs map quality test in view of sensed data of a target area, so as to effectively determine in time whether map data is qualified, thereby helping to improve safety of driving.

Description

Map quality detection method, computer-readable storage medium, and intelligent device

[0001] This application claims priority to Chinese patent application No. 202311727689.3 filed on December 14, 2023, entitled “A map quality detection method, computer-readable storage medium and intelligent device”. The entire contents of the above Chinese patent application are incorporated into this application by reference. Technical Field

[0002] The present application relates to the field of autonomous driving technology, and specifically provides a map quality detection method, a computer-readable storage medium, and an intelligent device. Background Art

[0003] Advanced driver assistance systems are gaining increasing attention. They primarily consist of the following modules: mapping, positioning, perception, environmental information, and planning and control. The positioning module provides the vehicle's specific location, the perception module identifies surrounding information (such as lane markings and obstacles), and the environmental information module integrates multi-source information (such as mapping and perception information) to output the vehicle's real-time environment, which the planning and control module uses to control the vehicle and make behavioral decisions.

[0004] Existing technologies primarily rely on high-precision maps for advanced driver assistance. However, due to limitations in mapping technology and costs, HD maps can contain various errors, such as geometric errors, semantic errors, and insufficient freshness. These errors, if not detected in a timely manner, can impact driving safety.

[0005] Summary of the Invention

[0006] This application aims to solve the above technical problems, namely, to solve the problem that existing high-precision maps may have errors, affecting the safety of the driving process.

[0007] In a first aspect, the present application provides a map quality detection method, which comprises:

[0008] Based on the map data of the target area, obtaining first attribute information of the map element to be detected;

[0009] acquiring second attribute information of a reference map element based on the perception data of the target area; the reference map element is associated with the map element to be detected;

[0010] comparing the first attribute information and the second attribute information;

[0011] A map quality detection result of the target area is determined at least according to the comparison result.

[0012] In some embodiments, acquiring the first attribute information of the map element to be detected includes: acquiring first semantic information of the map element to be detected; and acquiring the second attribute information of the reference map element includes: acquiring second semantic information of the reference map element.

[0013] In some embodiments, acquiring the first attribute information of the map element to be detected includes: acquiring first geometric information of the map element to be detected; and acquiring the second attribute information of the reference map element includes: acquiring second geometric information of the reference map element.

[0014] In some embodiments, determining the map quality detection result of the target area at least based on the comparison result includes:

[0015] When the comparison result is inconsistent, calculating the absolute value of the difference between the first geometric information and the second geometric information;

[0016] Comparing the absolute value of the difference with a preset threshold;

[0017] When the absolute value of the difference is greater than or equal to the preset threshold, it is determined that the map quality detection result is unqualified.

[0018] In some embodiments, acquiring the second attribute information of the reference map element based on the perception data of the target area includes:

[0019] Acquire multiple frames of the perception data; each frame of the perception data includes the second attribute information of at least one reference map element; or each frame of the perception data includes the second attribute information and identity information of at least one reference map element;

[0020] Associating and fusing the multiple frames of perception data to obtain fused perception data;

[0021] The second attribute information is obtained based on the fused perception data.

[0022] In some embodiments, associating the plurality of frames of the sensory data comprises:

[0023] When each frame of the perception data includes the identity information of at least one reference map element, for each reference map element, comparing the identity information of the reference map element in different frames of the perception data;

[0024] The plurality of reference map elements having the same identity information in different frames are associated with each other.

[0025] In some embodiments, before comparing the first attribute information and the second attribute information, the method further includes:

[0026] determining a confidence level of the reference map element based on the second attribute information;

[0027] comparing the confidence level to a confidence threshold;

[0028] When the confidence level is greater than or equal to the confidence level threshold, the step of comparing the first attribute information with the second attribute information is performed.

[0029] In some embodiments, before obtaining the second attribute information of the reference map element, the method further includes:

[0030] Obtaining the location information of the map element to be detected;

[0031] The reference map element is determined from the perception data based on the position information.

[0032] In a second aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the map quality detection method described in any one of the above items is implemented.

[0033] In a third aspect, the present application provides a smart device comprising:

[0034] at least one processor;

[0035] and, a memory communicatively coupled to the at least one processor;

[0036] The memory stores a computer program, and when the computer program is executed by the at least one processor, any one of the above-mentioned map quality detection methods is implemented.

[0037] By employing the above technical solution, the present application can obtain first attribute information of a map element to be detected based on the map data of the target area; obtain second attribute information of a reference map element based on the perception data of the target area; associate the reference map element with the map element to be detected; compare the first attribute information and the second attribute information; and determine a map quality detection result for the target area based on at least the comparison result. By combining the perception data of the target area for map quality detection, this method can promptly and effectively determine whether the map data is qualified, thereby improving driving safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The preferred embodiments of the present application are described below with reference to the accompanying drawings, in which:

[0039] Figures 1(1) to 1(3) are example diagrams of high-precision map errors provided by this application;

[0040] FIG2 is a flow chart showing the main steps of a map quality detection method provided in an embodiment of the present application;

[0041] FIG3 is a flowchart of a map quality detection method provided by a preferred embodiment of the present application;

[0042] FIG4 is a flowchart of a map quality detection method provided by another preferred embodiment of the present application;

[0043] FIG5 is a schematic diagram of the map quality detection system architecture provided in an embodiment of the present application;

[0044] FIG6 is a schematic diagram of a road scene provided by a specific example of this application;

[0045] FIG7 is a schematic diagram of the structure of a smart device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0046] Some embodiments of the present application are described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present application and are not intended to limit the scope of protection of the present application.

[0047] Based on the description of the background technology, it can be seen that due to factors such as mapping technology and mapping costs, high-precision maps may have various errors. See Figures 1(1)-1(3). The solid lines represent actual lanes and the dotted lines represent map lane lines. There may be map corner errors as shown in Figure 1(1), small curvature errors caused by insufficient mapping accuracy as shown in Figure 1(2), and map diversion errors caused by insufficient map freshness and untimely updates as shown in Figure 1(3).

[0048] Referring to FIG. 2 , FIG. 2 is a flow chart showing the main steps of a map quality detection method provided in an embodiment of the present application, which may include:

[0049] Step S11: acquiring first attribute information of a map element to be detected based on map data of a target area;

[0050] Step S12: Based on the perception data of the target area, obtaining second attribute information of the reference map element; associating the reference map element with the map element to be detected;

[0051] Step S13: Compare the first attribute information and the second attribute information;

[0052] Step S14: Determine the map quality detection result of the target area at least based on the comparison result.

[0053] In some embodiments, the map quality detection method provided in the present application can be applied to the driving scene of an intelligent driving device, and the target area can be the current driving environment area, and the target area can include multiple real map elements, such as at least one of lane lines, stop lines, traffic lights, speed limit signs and obstacles. The map data can be obtained based on the real scene of the target area, which may include at least one map element to be detected and at least one first attribute information of the map element to be detected, wherein the map element to be detected and its first attribute information may be consistent with the real map element and its attribute information in the target area; when the real scene of the target area changes and the map data is obtained based on the target area before the change, the map element to be detected and / or the first attribute information may deviate from the real scene. In some embodiments, the first attribute information may include first semantic information and / or first geometric information.

[0054] In some embodiments, the perception information may be the result of the output of an upstream perception module, which may be used to identify the image or point cloud data of the target area, and output each reference map element in the target area and the second attribute information and / or identity information corresponding to each reference map element, wherein the reference map element may correspond one-to-one to the real map element in the target area and the reference map element may be associated with the corresponding map element to be detected. In some embodiments, the reference map element may be obtained with the same position information as the map element to be detected as an association condition. When the map quality is qualified, the map element to be detected should correspond to the same real map element as the corresponding reference map element and the attribute information corresponding to each should also be the same as the attribute information of the real map element.

[0055] The second attribute information may include second semantic information and / or second geometric information.

[0056] In some embodiments, step S11 may specifically be to directly obtain first semantic information and / or first geometric information of the map element to be detected based on map data of the target area.

[0057] As an example, the first semantic information may include at least one of lane line virtuality and reality type, lane attributes, lane line color, and speed limit information. The first geometric information may include at least one of geometric dimension, curvature, position information, and curvature change rate.

[0058] In some embodiments, step S12 may specifically be to directly obtain the second semantic information and / or second geometric information of the reference map element based on the perception data of the target area.

[0059] As an example, the second semantic information may include at least one of lane line virtuality and reality type, lane attributes, lane line color, and speed limit information. The second geometric information may include at least one of geometric dimension, curvature, position information, and curvature change rate.

[0060] It should be noted that, in the embodiment of the present application, steps S11 and S12 may be executed simultaneously, or step S12 may be executed first and then S11. The above is merely an example of executing step S11 first and then S12, and there is no particular limitation on the execution order.

[0061] In some embodiments, when obtaining the first semantic information of the map element to be detected, the second semantic information of the reference map element may be correspondingly obtained; step S13 may specifically involve comparing the first semantic information with the second semantic information. The corresponding step S14 may specifically involve determining a map quality test result for the target area based on the comparison result. If the comparison result is consistent, the map quality test result for the target area is determined to be qualified; if the comparison result is inconsistent, the map quality test result for the target area is determined to be unqualified.

[0062] In other embodiments, when the first geometric information of the map element to be detected is obtained, the second geometric information of the reference map element can be obtained accordingly; step S13 can be specifically compared with the first geometric information and the second geometric information. Step S14 can be specifically determined based on the comparison result. When the comparison result is consistent, the map quality detection result of the target area is determined to be qualified; when the comparison result is inconsistent, the map quality detection result of the target area is determined to be unqualified. In other embodiments, step S14 can be specifically: when the comparison result is inconsistent, the absolute value of the difference between the first geometric information and the second geometric information is calculated; the absolute value of the difference is compared with a preset threshold; when the absolute value of the difference is greater than or equal to the preset threshold, the map quality detection result is determined to be unqualified. When the comparison result is inconsistent and the absolute value of the difference is less than the preset threshold, or when the comparison result is consistent, the map quality detection result is determined to be qualified. Compared with determining the map quality based only on the comparison results of the first geometric information and the second geometric information, judging by the absolute value of the difference between the first geometric information and the second geometric information combined with a preset threshold can effectively prevent misjudgment caused by inevitable errors, which is conducive to improving the accuracy of the map quality detection results.

[0063] In other embodiments, when obtaining the first semantic information and first geometric information of a map element to be detected and obtaining the second semantic information and second geometric information of a reference map element, step S13 may specifically involve comparing the first semantic information and the second semantic information, and comparing the first geometric information and the second geometric information. Accordingly, step S14 may specifically involve determining a map quality test result for the target area based on the semantic information comparison results and the geometric information comparison results. When the semantic information comparison results and the geometric information comparison results are consistent, the map quality test result for the target area is determined to be qualified; if the comparison results for either the semantic information or the geometric information are inconsistent, the map quality test result for the target area is determined to be unqualified. In other embodiments, step S14 may further specifically involve determining the map quality test result for the target area to be qualified when the semantic information comparison results and the geometric information comparison results are consistent, or when the semantic information comparison results are consistent and the geometric information comparison results are inconsistent, but the absolute value of the difference between the first and second geometric information is less than a preset threshold; otherwise, the map quality test result is determined to be unqualified. By testing map quality based on both semantic and geometric information, the reliability of the test results can be improved and its applicability can be expanded.

[0064] In some embodiments, single-frame perception data is prone to large noise. In order to obtain robust and accurate second attribute information, multiple frames of perception data can be spatiotemporally fused. For details, please refer to the description below.

[0065] Referring to FIG. 3 , FIG. 3 is a flowchart of a map quality detection method provided in a preferred embodiment of the present application, which may include:

[0066] Step S21: acquiring first attribute information of a map element to be detected based on map data of a target area;

[0067] Step S22: Acquire multiple frames of perception data; each frame of perception data includes second attribute information of at least one reference map element; or each frame of perception data includes second attribute information and identity information of at least one reference map element;

[0068] Step S23: Correlate and fuse the multiple frames of perception data to obtain fused perception data;

[0069] Step S24: obtaining second attribute information of the reference map element based on the fused perception data;

[0070] Step S25: Compare the first attribute information and the second attribute information;

[0071] Step S26: Determine the map quality detection result of the target area at least based on the comparison result.

[0072] Among them, steps S21, S25 and S26 can be implemented in the same manner as steps S11, S13 and S14, and are not described here for the sake of brevity. For details, please refer to the above description.

[0073] In some embodiments, step S22 may specifically involve acquiring multiple frames of perception data within a certain spatiotemporal range of the target area. The multiple frames of perception data may correspond to different acquisition times and / or acquisition locations. When corresponding to different acquisition locations, the multiple frames of perception data may also be converted to a unified coordinate system based on the real-time relative positioning information of the acquisition terminal. The reference map elements and the second attribute information may be configured in the same manner as in the embodiment corresponding to FIG2 .

[0074] In some embodiments, associating multiple frames of perception data in step S23 may specifically be:

[0075] When each frame of perception data includes identity information of at least one reference map element, for each reference map element, comparing the identity information of the reference map element in different frames of perception data;

[0076] Associate multiple reference map elements with consistent identity information in different frames.

[0077] The associated multiple reference map elements correspond to the same real map element.

[0078] In other embodiments, associating multiple frames of perception data in step S23 may specifically be:

[0079] Each frame of perception data includes the second geometric information of at least one reference map element, and the second geometric information can be position information. For any two adjacent frames of perception data, the perception data of the previous frame in the two adjacent frames of perception data is transformed into the coordinate system corresponding to the current frame of perception data through relative posture. Based on the converted perception data and any two reference map elements in the current frame of perception data, interval sampling is performed to obtain a set of sampling points corresponding to each reference map element, and the average distance between the two reference map elements is calculated based on the sampling point set. Using the Hungarian matching algorithm or the nearest neighbor matching algorithm, it is determined based on the average distance whether the two current reference map elements are associated. The same method can be used to obtain a reference map element association set for multiple frames of perception data. The reference map element association set can include multiple subsets, each subset representing a group of associated reference map elements. As an example, the reference map element association set can include a lane line association subset and a traffic light association subset.

[0080] In some embodiments, step S23 can also be used to model the reference map elements to be fused, construct a residual function using multiple reference map elements that have been successfully associated, and minimize the residual function using methods such as Gauss-Newton to obtain a fusion result, i.e., fused perception data. As an example, the reference map elements are lane lines, and multiple frames of perception data based on lane line association have been obtained. Construct a cubic curve model for the lane line: y = ax 3 +bx 2 +cx+d, (x, y) represents the position information of the lane line, and a, b, c, d represent the model parameters to be solved. Assuming that the model parameters of the cubic curve model of the lane line at the current moment have been obtained, for each lane line in the multi-frame lane lines associated with the current lane line: interval sampling is performed to obtain multiple sampling points (x1, y1), (x2, y2) ... (x n ,y n ), n is a positive integer, for each sampling point x i , using the cubic curve model we can get y i_hat =ax i 3 +bx i 2 +cx i +d, where y i_hat Represents the predicted y-axis coordinate of the i-th sampling point, x i Represents the x-axis coordinate of the i-th sampling point; set the residual function to loss = sum((y i -y i_hat ) 2 ), where sum represents the sum of the residuals at each sampling point to determine the overall lane line fit. The Gauss-Newton method is used to minimize the loss, resulting in optimized model parameters and, in turn, an optimized 3D lane line curve model. This optimized 3D lane line curve model serves as the fusion result. Fusion of multiple frames of correlated perception data facilitates obtaining more accurate and comprehensive secondary attribute information for reference map elements. It should be noted that other fusion methods known in the art may also be employed in other embodiments.

[0081] In some embodiments, step S24 may specifically be to obtain the second attribute information directly from the fused perception data.

[0082] The above is a map quality detection method provided by a preferred embodiment of the present application, which can achieve the same beneficial effects as the embodiment corresponding to Figure 2 above, and obtains fused perception data by acquiring multiple frames of perception data, associating and fusing the multiple frames of perception data; obtaining second attribute information based on the fused perception data is conducive to obtaining more accurate and comprehensive second attribute information of the reference map elements, thereby further improving the effectiveness of the map quality detection method.

[0083] In other embodiments, to ensure the effectiveness of the map quality detection method, the confidence level of the second attribute information can be pre-determined, so that map quality can be detected based on the second attribute information with a high confidence level. For details, please refer to the following embodiments. It should be noted that the following embodiments can be implemented based on the embodiment corresponding to Figure 2 or Figure 3 . The following description uses the embodiment based on Figure 3 as an example.

[0084] Referring to FIG. 4 , FIG. 4 is a flowchart of a map quality detection method provided by another preferred embodiment of the present application, which may include:

[0085] Step S31: obtaining first attribute information of a map element to be detected based on map data of a target area;

[0086] Step S32: Acquire multiple frames of perception data; each frame of perception data includes second attribute information of at least one reference map element; or each frame of perception data includes second attribute information and identity information of at least one reference map element;

[0087] Step S33: Associating and fusing the multiple frames of perception data to obtain fused perception data;

[0088] Step S34: obtaining second attribute information based on the fused perception data;

[0089] Step S35: determining the confidence level of the reference map element based on the second attribute information;

[0090] Step S36: comparing the confidence level with the confidence level threshold;

[0091] When the confidence level is greater than or equal to the confidence threshold, execute steps S37 and S38; when the confidence level is less than the confidence threshold, end the process.

[0092] Step S37: Compare the first attribute information and the second attribute information;

[0093] Step S38: Determine the map quality detection result of the target area at least based on the comparison result.

[0094] Among them, steps S31-S34, S37 and S38 can be implemented in the same manner as steps S21-S26. For the sake of brevity, they are not repeated here. Please refer to the above description for details.

[0095] In some embodiments, step S35 may specifically involve comparing the second attribute information with pre-annotated information corresponding to the reference map element and determining a confidence level based on the comparison result. In other embodiments, the confidence level of the current second attribute information may be determined based on the type of the second attribute information and a preset relationship table. For example, in complex scenarios such as intersections and diverging / merging traffic, the second attribute information derived from the perception data may contain significant errors, and a lower confidence level may be set accordingly.

[0096] In some embodiments, the confidence threshold can be set as needed.

[0097] By obtaining the confidence level of the second attribute information and comparing the confidence level with the confidence threshold, the validity of the second attribute information can be determined, thereby ensuring the validity of the map quality detection result.

[0098] In some embodiments, before obtaining the second attribute information of the reference map element, the map quality detection method provided by this application may further include:

[0099] Get the location information of the map element to be detected;

[0100] A reference map element is determined from the perception data based on the position information.

[0101] In some embodiments, the location information of the map element to be detected may be obtained based on map data.

[0102] In some embodiments, the determined positions of the reference map element and the map element to be detected are the same, or the distances between them are within a certain range.

[0103] In another aspect of the present application, a map quality detection system is provided. FIG. 5 is a schematic diagram of the map quality detection system architecture provided in an embodiment of the present application. The system architecture can be applied to a road scene of a driving device to implement the map quality detection method of the embodiment corresponding to FIG. 4 , which can include:

[0104] The vehicle-side perception data acquisition module is used to obtain multi-frame perception data of the current road scene;

[0105] Vehicle-side relative positioning module, used to obtain vehicle-side relative positioning information in real time;

[0106] The perception spatiotemporal fusion module is used to associate and fuse multiple frames of perception data to obtain fused perception data; and based on the acquired vehicle-side relative positioning information, the fused perception data is converted into the vehicle-mounted coordinate system of the current vehicle;

[0107] High-precision map acquisition module, used to obtain high-precision maps;

[0108] The vehicle-side global positioning module is used to obtain the global positioning information of the current vehicle;

[0109] The map data acquisition module is used to extract local high-precision map data of the vehicle's current road scene from the high-precision map based on the vehicle's global positioning information;

[0110] The perception map association module is used to use the spatial location information of each map element and associate the map elements in the fused perception data and local high-precision map data based on the nearest neighbor matching algorithm;

[0111] The confidence calculation module is used to calculate the confidence of the fused perception data and compare the confidence with the confidence threshold. When the confidence is greater than or equal to the confidence threshold, it is determined that the confidence is high and enters the quality inspection module; when the confidence is less than the confidence threshold, the process ends;

[0112] The quality inspection module may include a geometric quality inspection submodule and a semantic quality inspection submodule. The geometric quality inspection submodule is used to compare the geometric information of the reference map elements in the fused perception data and the map elements to be detected in the local high-precision map data that are associated with the reference map elements, and output the quality inspection results based on the comparison results; alternatively, the absolute value of the difference may be calculated based on the comparison results, and the absolute value of the difference may be compared with a preset threshold, and the quality inspection results may be output according to the size relationship between the absolute value of the difference and the preset threshold; the semantic quality inspection submodule is used to compare the semantic information of the reference map elements in the fused perception data and the map elements to be detected in the local high-precision map data that are associated with the reference map elements, and output the quality inspection results according to the comparison results.

[0113] The map quality detection system provided in this application can achieve the same beneficial effects as the corresponding embodiment of FIG4 when it is in operation.

[0114] See Figure 6, which is a schematic diagram of a road scene provided by a specific example of this application, in which construction occurs in front of the vehicle, the original road is blocked, and a new road is opened next to it. If the high-precision map is not updated in time, a map error of the map diversion type shown in Figure 1 (3) will occur. Perception data can reflect real-time road conditions. The map quality detection method provided by this application can detect map anomalies in advance and output abnormal signals to the downstream. So that after the downstream receives the map anomaly signal, it abandons the use of the map to construct local environment information and replaces it with the use of perception data, ensuring the correctness of the local environment on the vehicle side, so that the vehicle can drive according to the lane line output based on the perception data, thereby avoiding safety problems. The specific road quality detection process may include:

[0115] Step 1: Obtain fused perception data: Obtain multi-frame perception data within a 20m range behind the ego vehicle. Use the ego vehicle relative positioning information corresponding to each frame of perception data to convert all multi-frame perception data into the ego vehicle coordinate system. Use the identity information of the perceived lane lines to complete the preliminary association of the multi-frame perception data. For the preliminarily associated perception lane line matching pairs, calculate the average distance between the two perception lane line point clouds. If the average distance is greater than the distance threshold, it is determined to be a false match. If it is less than or equal to the distance threshold, it is determined to be a successful match, and the current perception lane line matching pair is determined to be associated.

[0116] Step 2: Obtain a local high-precision map: Based on the global positioning information, capture the high-precision map information of the vehicle's current road scene and convert it to the vehicle's coordinate system.

[0117] Step 3: Perception Map Association: The vehicle coordinate system is constructed using the right-hand rule, with the x-axis oriented in the vehicle's forward direction and the y-axis oriented to the left. For each lane line in the perception data and the local HD map, the y-value at x = 0 is calculated, denoted as C0. Left and right lane lines are distinguished based on C0. Specifically, among all lane lines with C0 < 0, the lane line with the largest C0 is designated as the left lane line; among all lane lines with C0 > 0, the lane line with the smallest C0 is designated as the right lane line. This allows for the correspondence between the perception data and the local HD map for both the left and right lane lines.

[0118] Step 4: Calculate the confidence of the perception data: Determine the confidence of the perception data based on the type of the current road scene.

[0119] Step 5: Geometric quality inspection: The left and right lane lines are inspected separately. Specifically, the lane lines are divided into segments with a step size of x = 10m. The average distance Dn between the lane line corresponding to the perception data and the lane line corresponding to the local high-precision map in each segment is calculated. If the average distance Dn of a segment is greater than the distance threshold of 1m, it is determined that a geometric error exists and the output map quality inspection result is unqualified, so that downstream can make timely adjustments based on this to avoid safety issues.

[0120] It will be understood by those skilled in the art that all or part of the processes in the method for implementing the above-mentioned embodiments of the present application can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of the above-mentioned various method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. The computer-readable storage medium can include: any entity or device, medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory, random access memory, electric carrier signal, telecommunication signal and software distribution medium, etc. that can carry the computer program code.

[0121] In another aspect, the present application provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the map quality detection method described in any of the above embodiments. The computer-readable storage medium may be a storage device formed by various electronic devices. Optionally, in embodiments of the present application, the computer-readable storage medium is non-transitory.

[0122] Another aspect of the present application provides an intelligent device, which may include at least one processor; and a memory communicatively connected to the at least one processor; wherein a computer program is stored in the memory, and when the computer program is executed by the at least one processor, the map quality detection method described in any of the above embodiments is implemented.

[0123] The smart devices described in this application may include driving devices, smart cars, robots and other devices.

[0124] Referring to FIG. 7 , FIG. 7 exemplarily shows a structure in which a memory 71 and a processor 72 are connected via a bus, and only one memory 71 and only one processor 72 are provided.

[0125] In other embodiments, the smart device may include multiple memories 71 and multiple processors 72. The program for executing the map quality detection method of any of the above-described embodiments may be divided into multiple subroutines, each of which may be loaded and executed by the processor 72 to perform different steps of the map quality detection method of the above-described method embodiment. Specifically, each subroutine may be stored in a different memory 71, and each processor 72 may be configured to execute the programs in one or more memories 71 to collectively implement the map quality detection method of the above-described method embodiment.

[0126] In some embodiments of the present application, the smart device further includes at least one sensor configured to sense information. The at least one sensor is communicatively coupled to any of the processors described herein. Optionally, the smart device further includes an autonomous driving system configured to guide the smart device to autonomously drive or provide assisted driving. The at least one processor communicates with the at least one sensor and / or the autonomous driving system to implement the map quality detection method described in any of the aforementioned embodiments of the present application.

[0127] The relevant user personal information that may be involved in the various embodiments of this application is strictly in accordance with the requirements of laws and regulations, following the principles of legality, legitimacy and necessity, and based on the reasonable purposes of business scenarios, to process the personal information that users actively provide during the use of products / services or generated due to the use of products / services, as well as the personal information obtained with the user's authorization.

[0128] The user personal information processed by this application will vary depending on the specific product / service scenario and must be based on the specific scenario in which the user uses the product / service. This may involve the user's account information, device information, driving information, vehicle information, or other related information. This application will treat the user's personal information and its processing with a high degree of diligence.

[0129] This application attaches great importance to the security of user personal information and has taken reasonable and feasible security protection measures that comply with industry standards to protect user information and prevent personal information from being accessed, disclosed, used, modified, damaged or lost without authorization.

[0130] Thus far, the technical solutions of the present application have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is readily understood by those skilled in the art that the scope of protection of the present application is obviously not limited to these specific embodiments. Without departing from the principles of the present application, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present application.

Claims

1. A method for detecting map quality, characterized in that, Including: Based on the map data of the target area, obtaining first attribute information of the map element to be detected; Based on the perception data of the target area, obtaining second attribute information of a reference map element; the reference map element is associated with the map element to be detected; Comparing the first attribute information and the second attribute information; Determining at least according to the comparison result a map quality detection result of the target area.

2. The method according to claim 1, characterized in that, The obtaining of the first attribute information of the map element to be detected includes: obtaining first semantic information of the map element to be detected; the obtaining of the second attribute information of the reference map element includes: obtaining second semantic information of the reference map element.

3. The method according to claim 1, characterized in that, The obtaining of the first attribute information of the map element to be detected includes: obtaining first geometric information of the map element to be detected; the obtaining of the second attribute information of the reference map element includes: obtaining second geometric information of the reference map element.

4. The method according to claim 3, characterized in that, The determining at least according to the comparison result a map quality detection result of the target area includes: When the comparison result is inconsistent, calculating an absolute value of a difference between the first geometric information and the second geometric information; Comparing the absolute value of the difference with a preset threshold; When the absolute value of the difference is greater than or equal to the preset threshold, determining that the map quality detection result is unqualified.

5. The method according to any one of claims 1 to 4, characterized in that, The obtaining of the second attribute information of the reference map element based on the perception data of the target area includes: Obtaining multiple frames of the perception data; each frame of the perception data includes at least one piece of the second attribute information of the reference map element; or, each frame of the perception data includes at least one piece of the second attribute information of the reference map element and identity identification information; Associating and fusing the multiple frames of the perception data to obtain fused perception data; Obtaining the second attribute information based on the fused perception data.

6. The method according to claim 5, characterized in that, The associating of the multiple frames of the perception data includes: When each frame of the perception data includes at least one piece of the identity identification information of the reference map element, for each reference map element, comparing the identity identification information of the reference map element in different frames of the perception data; Associating multiple reference map elements with consistent identity identification information in different frames.

7. The method according to any one of claims 1 to 4, characterized in that, Before the comparing of the first attribute information and the second attribute information, the method further includes: Determining a confidence level of the reference map element based on the second attribute information; Comparing the confidence level with a confidence level threshold; When the confidence level is greater than or equal to the confidence level threshold, then performing the step of comparing the first attribute information and the second attribute information.

8. The method according to any one of claims 1 to 4, characterized in that, Before the obtaining of the second attribute information of the reference map element, the method further includes: Obtaining position information of the map element to be detected; Determining the reference map element from the perception data based on the position information.

9. A computer-readable storage medium, characterized in that, A computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the map quality detection method according to any one of claims 1 to 8 is implemented.

10. An intelligent device, characterized in that, Including: At least one processor; And a memory communicatively connected to the at least one processor; Wherein, a computer program is stored in the memory, and when the computer program is executed by the at least one processor, the map quality detection method according to any one of claims 1 to 8 is implemented.

Images