An electronic navigation map quality detection method and device

Abstract

Embodiments of the present application disclose an electronic navigation map quality detection method and device, the method comprising: obtaining an electronic navigation map to be detected; obtaining a plurality of pose information and corresponding positioning time thereof in a driving process of a vehicle in a scene corresponding to the electronic navigation map; obtaining perception data determined by the vehicle in the driving process, a collection time corresponding to each road image being in a corresponding relationship with the positioning time corresponding to each pose information; for each pose information, determining map semantic data corresponding to the pose information from the electronic navigation map; and for each pose information, determining a quality detection result of the electronic navigation map based on map position information of the map semantic data corresponding to the pose information in the electronic navigation map and / or observation position information of the perception data corresponding to the pose information in the corresponding road image, so as to realize detection of quality problems of the electronic navigation map.

Description

[0001] This application is a divisional application of application No. 201910778638.0, entitled "Electronic navigation map quality detection method and device", and the parent application has an application date of August 22, 2019. TECHNICAL FIELD

[0002] The present application relates to the technical field of intelligent transportation, in particular to an electronic navigation map quality detection method and device. BACKGROUND

[0003] In the field of unmanned driving, the decision and planning technology of the driving route of a vehicle often relies on the electronic navigation map used by the vehicle. Especially for the positioning of an unmanned vehicle, the positional accuracy of the electronic navigation map determines the accuracy of the positioning result of the unmanned vehicle to a certain extent.

[0004] In related technologies, during the creation of an electronic navigation map, quality problems often occur in the electronic navigation map, such as bending, breaking of lane lines, shaking of lane line elevations, and tilting of street lamp poles, etc. The above quality problems affect the accuracy of the positioning result of the vehicle during driving to a certain extent.

[0005] Therefore, the detection of the quality problems of the electronic navigation map is a key problem for obtaining a high-quality electronic navigation map. SUMMARY

[0006] The present application provides an electronic navigation map quality detection method and device to realize the detection of quality problems of an electronic navigation map. The specific technical solutions are as follows:

[0007] In a first aspect, the present application provides an electronic navigation map quality detection method, comprising:

[0008] obtaining an electronic navigation map to be detected;

[0009] obtaining trajectory information of a vehicle during driving in a scene corresponding to the electronic navigation map, wherein the trajectory information comprises a plurality of pose information and a positioning time corresponding to each pose information;

[0010] obtaining perception data determined by the vehicle during driving, wherein each perception data is data detected from a road image collected by an image collection device of the vehicle, and the collection time corresponding to each road image and the positioning time corresponding to each pose information have a corresponding relationship;

[0011] for each pose information, determining map semantic data corresponding to the pose information from the electronic navigation map;

[0012] For each pose information, based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map, and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image, a quality detection result of the electronic navigation map is determined.

[0013] Optionally, the quality detection result includes a detection result of a position deviation condition of each map semantic data.

[0014] The step of determining, for each pose information, a quality detection result of the electronic navigation map based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map, and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image, includes:

[0015] For each pose information, the projection position information of the map semantic data in the road image corresponding to the pose information is determined based on the map position information of the map semantic data corresponding to the pose information and the pose information.

[0016] Based on the projection position information of the map semantic data in the road image corresponding to the pose information and the observation position information of the perception data corresponding to the map semantic data in the road image, a position deviation value between the map semantic data and the perception data corresponding thereto is determined.

[0017] It is judged whether the position deviation value exceeds a preset distance threshold.

[0018] If the judgment result is that the position deviation value exceeds the preset distance threshold, it is determined that there is a position deviation condition of the map semantic data in the electronic navigation map.

[0019] Optionally, the perception data includes a perceived lane line, and the map semantic data includes a map lane line; the quality detection result includes a detection result of a position deviation of the map lane line in the elevation direction.

[0020] The step of determining, for each pose information, a quality detection result of the electronic navigation map based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map, and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image, includes:

[0021] For each pose information, the mapping position information of the map lane line in the body coordinate system corresponding to the pose information is determined based on the map position information of the map lane line in the map semantic data corresponding to the pose information and the pose information.

[0022] For each pose information, based on observation position information of a perception lane line in the perception data corresponding to the pose information in a corresponding road image, the pose information, and a projection matrix corresponding to the image acquisition device, mapping position information of the perception lane line in a vehicle body coordinate system corresponding to the pose information is determined;

[0023] For each map lane line corresponding to each pose information, based on mapping position information of the map lane line in a vehicle body coordinate system corresponding to the pose information, and mapping position information of a perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information, whether there is a position deviation situation of the map lane line in a horizontal axis direction and a vertical axis direction of the vehicle body coordinate system is determined;

[0024] If it is determined that there is no position deviation situation of the map lane line in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system, for each pose information, based on map position information of a map lane line in the electronic navigation map in map semantic data corresponding to the pose information, and the pose information, projection position information of the map lane line in a road image corresponding to the pose information is determined.

[0025] For each map lane line corresponding to each pose information, based on projection position information of the map lane line in a road image corresponding to the pose information, and observation position information of a perception lane line corresponding to the map lane line in the road image, whether there is a position deviation situation of the map lane line in an elevation direction is determined.

[0026] Optionally, the perception data includes a perception lane line, and the map semantic data includes a map lane line; the quality detection result includes a detection result of a bending condition and / or a broken condition of the map lane line;

[0027] If the quality detection result includes a detection result of a bending condition of the map lane line, the step of determining, for each pose information, a quality detection result of the electronic navigation map based on map position information of a map lane line in the electronic navigation map in map semantic data corresponding to the pose information, and / or observation position information of a perception lane line in a corresponding road image corresponding to the pose information, includes:

[0028] For each pose information, based on map position information of a map lane line in the electronic navigation map in map semantic data corresponding to the pose information, and the pose information, mapping position information of the map lane line in a vehicle body coordinate system corresponding to the pose information is determined, wherein each mapping position information includes mapping position information of a plurality of discrete points corresponding to the corresponding map lane line;

[0029] For each map lane line corresponding to each pose information, a lane line fitting line corresponding to the map lane line is fitted based on mapping position information of a plurality of discrete points corresponding to the map lane line and a preset fitting algorithm;

[0030] For each map lane line corresponding to each pose information, a distance variance corresponding to the map lane line is determined based on the lane line fitting line corresponding to the map lane line and mapping position information of a plurality of discrete points corresponding to the map lane line;

[0031] If the distance variance exceeds a preset variance threshold, it is determined that the map lane line has a lane line bending condition;

[0032] And / or, if the quality detection result includes a detection result of a map lane line breakage condition, the step of determining, for each pose information, a quality detection result of the electronic navigation map based on map position information of map semantic data corresponding to the pose information in the electronic navigation map and / or observation position information of perception data corresponding to the pose information in the corresponding road image, comprises:

[0033] For each pose information, conversion position information of a map lane line in the map semantic data corresponding to the pose information in the electronic navigation map and in the pose information is determined based on the map lane line, wherein each conversion position information comprises conversion position information of a plurality of discrete points corresponding to the corresponding map lane line;

[0034] For each map lane line corresponding to each pose information, a distance between each adjacent two discrete points is calculated based on conversion position information of a plurality of discrete points corresponding to the map lane line;

[0035] If, among the plurality of discrete points corresponding to the map lane line, there are two adjacent discrete points whose distance exceeds a preset difference value from the distance between other adjacent two discrete points, it is determined that the map lane line has a lane line breakage condition.

[0036] Optionally, the perception data includes a perceived street lamp pole, and the map semantic data includes a map street lamp pole; and the quality detection result includes a detection result of a map street lamp pole tilting condition.

[0037] The step of determining, for each pose information, a quality detection result of the electronic navigation map based on map position information of map semantic data corresponding to the pose information in the electronic navigation map and / or observation position information of perception data corresponding to the pose information in the corresponding road image, comprises:

[0038] For each pose information, based on the map position information of the map road lamp pole in the map semantic data corresponding to the pose information and the pose information, determine the projection position information of the map road lamp pole in the road image corresponding to the pose information;

[0039] For each map road lamp pole corresponding to each pose information, based on the projection position information of the map road lamp pole in the road image corresponding to the pose information and the observation position information of the perception road lamp pole in the corresponding road image in the perception data corresponding to the map road lamp pole, determine whether the map road lamp pole has an inclination condition.

[0040] Optionally, the perception data includes a perception traffic sign, and the map semantic data includes a map traffic sign; and the quality detection result includes a detection result of a map traffic sign overturning condition.

[0041] The step of determining, for each pose information, the quality detection result of the electronic navigation map based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image, comprises:

[0042] For each pose information, based on the map position information of the map traffic sign in the map semantic data corresponding to the pose information in the electronic navigation map and the pose information, determine the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information.

[0043] For each pose information, based on the observation position information of the perception traffic sign in the corresponding road image in the perception data corresponding to the pose information, the pose information and the projection matrix corresponding to the image acquisition device, determine the mapping position information of the perception traffic sign in the vehicle body coordinate system corresponding to the pose information.

[0044] For each map traffic sign corresponding to each pose information, based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information, determine whether the map traffic sign has an overturning condition.

[0045] Optionally, the step of determining, based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information, whether the map traffic sign has an overturning condition, comprises:

[0046] Based on the mapping position information of the traffic sign in the vehicle coordinate system corresponding to the pose information and a preset plane fitting algorithm, a first fitting plane corresponding to the traffic sign in the map is fitted;

[0047] Based on the mapping position information of the perceived traffic sign corresponding to the traffic sign in the map in the vehicle coordinate system corresponding to the pose information and the preset plane fitting algorithm, a second fitting plane corresponding to the perceived traffic sign corresponding to the traffic sign in the map is fitted;

[0048] An included angle between a normal vector corresponding to the first fitting plane and a normal vector corresponding to the second fitting plane is calculated;

[0049] If the included angle exceeds a preset angle, it is determined that the traffic sign in the map has a flipping situation.

[0050] Optionally, after the step of determining the quality detection result of the electronic navigation map based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image for each pose information, the method further comprises:

[0051] If the quality detection result of the electronic navigation map indicates that the electronic navigation map has a quality problem, an abnormal statistical document and abnormal image information corresponding to the electronic navigation map are stored, wherein the abnormal statistical document at least includes an identifier of the map semantic data having a quality problem in the electronic navigation map and a problem quality type thereof, map position information in the electronic navigation map, and the abnormal image information includes image information corresponding to the map semantic data having a quality problem.

[0052] In a second aspect, an embodiment of the present application provides an electronic navigation map quality detection device, comprising:

[0053] A first obtaining module configured to obtain an electronic navigation map to be detected;

[0054] A second obtaining module configured to obtain trajectory information of a vehicle in a scene corresponding to the electronic navigation map during driving, wherein the trajectory information includes a plurality of pose information and a positioning time corresponding to each pose information;

[0055] A third obtaining module configured to obtain perception data determined by the vehicle during the driving, wherein each perception data is data detected from a road image collected by an image collection device of the vehicle, and a collection time corresponding to each road image has a corresponding relationship with the positioning time corresponding to each pose information;

[0056] The first determining module is configured to determine, for each piece of pose information, map semantic data corresponding to the pose information from the electronic navigation map;

[0057] The second determining module is configured to determine, for each piece of pose information, a quality detection result of the electronic navigation map based on map position information of the map semantic data corresponding to the pose information in the electronic navigation map and / or observation position information of the perception data corresponding to the pose information in the corresponding road image.

[0058] Optionally, the quality detection result includes a detection result of a position deviation condition of each piece of map semantic data.

[0059] The second determining module is specifically configured to determine, for each piece of pose information, projection position information of the map semantic data in a road image corresponding to the pose information based on map position information of the map semantic data in the electronic navigation map and the pose information.

[0060] The position deviation value between the map semantic data and the perception data corresponding to the map semantic data is determined based on the projection position information of the map semantic data in the road image corresponding to the pose information and the observation position information of the perception data corresponding to the map semantic data in the road image.

[0061] It is determined whether the position deviation value exceeds a preset distance threshold.

[0062] If the determination result is that the position deviation value exceeds the preset distance threshold, it is determined that there is a position deviation condition of the map semantic data in the electronic navigation map.

[0063] Optionally, the perception data includes a perception lane line, and the map semantic data includes a map lane line; and the quality detection result includes a detection result of a position deviation of the map lane line in an elevation direction.

[0064] The second determining module is specifically configured to determine, for each piece of pose information, mapping position information of the map lane line in a vehicle coordinate system corresponding to the pose information based on the map position information of the map lane line in the electronic navigation map and the pose information.

[0065] For each piece of pose information, mapping position information of a perception lane line in a vehicle coordinate system corresponding to the pose information is determined based on observation position information of the perception lane line in a corresponding road image, the pose information, and a projection matrix corresponding to the image acquisition device.

[0066] For each map lane corresponding to each pose information, based on the mapping position information of the map lane in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception lane corresponding to the map lane in the vehicle body coordinate system corresponding to the pose information, it is determined whether there is a position deviation of the map lane in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system;

[0067] If it is determined that there is no position deviation of the map lane in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system, for each pose information, based on the map position information of the map lane in the electronic navigation map in the map semantic data corresponding to the pose information and the pose information, the projection position information of the map lane in the road image corresponding to the pose information is determined.

[0068] For each map lane corresponding to each pose information, based on the projection position information of the map lane in the road image corresponding to the pose information and the observation position information of the perception lane in the road image in the perception data corresponding to the map lane, it is determined whether there is a position deviation of the map lane in the elevation direction.

[0069] Optionally, the perception data includes a perception lane, and the map semantic data includes a map lane; the quality detection result includes a detection result of a bending condition and / or a broken condition of the map lane;

[0070] If the quality detection result includes a detection result of a bending condition of the map lane, the second determination module is specifically configured to, for each pose information, based on the map position information of the map lane in the electronic navigation map in the map semantic data corresponding to the pose information and the pose information, determine the mapping position information of the map lane in the vehicle body coordinate system corresponding to the pose information, wherein each mapping position information includes mapping position information of a plurality of discrete points corresponding to the corresponding map lane.

[0071] For each map lane corresponding to each pose information, based on the mapping position information of a plurality of discrete points corresponding to the map lane and a preset fitting algorithm, a lane fitting line corresponding to the map lane is fitted;

[0072] For each map lane corresponding to each pose information, based on the lane fitting line corresponding to the map lane and the mapping position information of a plurality of discrete points corresponding to the map lane, a distance variance corresponding to the map lane is determined.

[0073] If the distance variance exceeds a preset variance threshold, it is determined that the map lane has a lane bending condition;

[0074] And / or, if the quality detection result includes a detection result of a map lane line breakage case, the second determination module is specifically configured to, for each pose information, determine, based on map lane line in map position information of the electronic navigation map and the pose information, a conversion position information of the map lane line in a vehicle body coordinate system corresponding to the pose information, where each conversion position information includes conversion position information of a plurality of discrete points corresponding to the corresponding map lane line.

[0075] For each map lane line corresponding to each pose information, a distance between each two adjacent discrete points is calculated based on the conversion position information of the plurality of discrete points corresponding to the map lane line.

[0076] If the distance between the two adjacent discrete points among the plurality of discrete points corresponding to the map lane line exceeds a preset difference value from the distance between other two adjacent discrete points, it is determined that the map lane line has a lane line breakage case.

[0077] Optionally, the perception data includes a perception street lamp pole, and the map semantic data includes a map street lamp pole; the quality detection result includes a detection result of a map street lamp pole tilting condition.

[0078] The second determination module is specifically configured to, for each pose information, determine, based on map street lamp pole in map position information of the electronic navigation map and the pose information, a projection position information of the map street lamp pole in a road image corresponding to the pose information.

[0079] For each map street lamp pole corresponding to each pose information, it is determined whether the map street lamp pole has a tilting condition based on the projection position information of the map street lamp pole in the road image corresponding to the pose information and observation position information of the perception street lamp pole in the corresponding road image in the perception data corresponding to the map street lamp pole.

[0080] Optionally, the perception data includes a perception traffic sign, and the map semantic data includes a map traffic sign; the quality detection result includes a detection result of a map traffic sign overturning case.

[0081] The second determination module is specifically configured to, for each pose information, determine, based on map traffic sign in map position information of the electronic navigation map and the pose information, a mapping position information of the map traffic sign in a vehicle body coordinate system corresponding to the pose information.

[0082] For each pose information, based on observation position information of a perceived traffic sign in a corresponding road image in perception data corresponding to the pose information, the pose information, and a projection matrix corresponding to the image collection device, mapping position information of the perceived traffic sign in a vehicle body coordinate system corresponding to the pose information is determined.

[0083] For a map traffic sign corresponding to each pose information, based on mapping position information of the map traffic sign in a vehicle body coordinate system corresponding to the pose information, and mapping position information of a perceived traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information, whether the map traffic sign has a flip situation is determined.

[0084] Optionally, the second determination module is specifically configured to fit a first fitting surface corresponding to the map traffic sign based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information and a preset plane fitting algorithm.

[0085] fit a second fitting surface corresponding to the perceived traffic sign corresponding to the map traffic sign based on the mapping position information of the perceived traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information and the preset plane fitting algorithm.

[0086] An angle between a normal vector corresponding to the first fitting surface and a normal vector corresponding to the second fitting surface is calculated.

[0087] If the angle exceeds a preset angle, it is determined that the map traffic sign has a flip situation.

[0088] Optionally, the device further comprises:

[0089] The storage module is configured to, after the quality detection result of the electronic navigation map is determined based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map, and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image, if the quality detection result of the electronic navigation map indicates that the electronic navigation map has a quality problem, store an abnormal statistical document and abnormal image information corresponding to the electronic navigation map, wherein the abnormal statistical document at least includes an identifier of the map semantic data having a quality problem in the electronic navigation map and a problem quality type thereof, map position information in the electronic navigation map, and the abnormal image information includes image information corresponding to the map semantic data having a quality problem.

[0090] From the above, the electronic navigation map quality detection method and device provided by the embodiment of the application can obtain an electronic navigation map to be detected, obtain trajectory information of a vehicle in a scene corresponding to the electronic navigation map during driving, wherein the trajectory information comprises a plurality of pose information and positioning time corresponding to each pose information, obtain perception data determined by the vehicle during driving, wherein each perception data is data recognized from a road image collected by an image collection device of the vehicle, and a collection time corresponding to each road image and a positioning time corresponding to each pose information have a corresponding relationship, for each pose information, determine map semantic data corresponding to the pose information from the electronic navigation map, and for each pose information, determine a quality detection result of the electronic navigation map based on map position information of the map semantic data corresponding to the pose information in the electronic navigation map and / or observation position information of the perception data corresponding to the pose information in the corresponding road image.

[0091] By applying the embodiment of the application, the map semantic data and the perception data corresponding to each other in the electronic navigation map can be determined based on the positioning time corresponding to the pose information in the trajectory information of the vehicle in the scene corresponding to the electronic navigation map during driving, and then, for each pose information, the quality detection result of the electronic navigation map is determined based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image, so as to realize the detection of the quality problem of the electronic navigation map. Of course, any product or method implemented by the application does not necessarily need to achieve all the advantages described above.

[0092] The innovation points of the embodiment of the application include:

[0093] 1. The map semantic data and the perception data corresponding to each other in the electronic navigation map can be determined based on the positioning time corresponding to the pose information in the trajectory information of the vehicle in the scene corresponding to the electronic navigation map during driving, and then, for each pose information, the quality detection result of the electronic navigation map is determined based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image, so as to realize the detection of the quality problem of the electronic navigation map.

[0094] 2. Based on the projection position information of each map semantic data in the corresponding image in the electronic navigation map and the observation position information of the corresponding perception data in the road image, whether the position of each map semantic data in the electronic navigation map is deviated or not can be determined.

[0095] 3、If the perception data contains the perception lane line, the map lane line is mapped from the electronic navigation map to the vehicle body coordinate system, and the perception lane line corresponding to the map lane line is mapped from the road image to the vehicle body coordinate system, and then, based on the mapping position information of the map lane line in the vehicle body coordinate system and the mapping position information of the perception lane line corresponding to the map lane line in the vehicle body coordinate system, it is determined whether the map lane line appears position deviation from different angles, so as to realize the detection of the position deviation of the map lane line in the electronic navigation map.

[0096] 4、If the perception data contains the perception lane line, the map lane line is mapped from the electronic navigation map to the vehicle body coordinate system, and based on the mapping position information of the plurality of discrete points corresponding to the map lane line, the lane line fitting line corresponding to the map lane line is fitted, and based on the mapping position information of the plurality of discrete points corresponding to the map lane line and the lane line fitting line, the distance variance corresponding to the map lane line is determined, and based on the distance variance, it is determined whether the map lane line has the lane line bending condition, and / or based on the conversion position information of the plurality of discrete points corresponding to the map lane line in the vehicle body coordinate system or in the road image, the distance between each adjacent two discrete points is calculated, and based on the distance between each adjacent two discrete points, it is determined whether the map lane line has the lane line breakage condition, so as to realize the detection of the lane line bending condition and / or the lane line breakage condition of the map lane line in the electronic navigation map.

[0097] 5、If the perception data contains the perception street lamp pole, the map street lamp pole is projected into the image corresponding to the perception street lamp pole corresponding to the map street lamp pole, and based on the projection position information of the map street lamp pole in the road image and the observation position information of the perception street lamp pole corresponding to the map street lamp pole in the road image, the detection of the inclination condition of the map street lamp pole in the electronic navigation map is realized.

[0098] 6、If the perception data contains the perception traffic sign, the map traffic sign in the map semantic data is mapped to the vehicle body coordinate system, and the perception traffic sign corresponding to the map traffic sign is mapped to the vehicle body coordinate system, and then, based on the mapping position information of the map traffic sign in the vehicle body coordinate system and the mapping position information of the perception traffic sign corresponding to the map traffic sign in the vehicle body coordinate system, the detection of the overturning condition of the map traffic sign in the electronic navigation map is realized. BRIEF DESCRIPTION OF DRAWINGS

[0099] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application. Those skilled in the art can also obtain other drawings according to these drawings without any creative effort.

[0100] Figure 1 A flowchart of an electronic navigation map quality detection method provided by an embodiment of the present application is shown in FIG. 1.

[0101] Figure 2 Another flowchart of an electronic navigation map quality detection method provided by an embodiment of the present application is shown in FIG. 2.

[0102] Figure 3 A structural diagram of an electronic navigation map quality detection device provided by an embodiment of the present application is shown in FIG. 3. DETAILED DESCRIPTION

[0103] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort belong to the scope of protection of the present application.

[0104] It should be noted that the terms "include" and "have" and any variations thereof in the embodiments of the present application and the drawings are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but can optionally include other steps or units not listed or can optionally include other steps or units inherent to the process, method, product or device.

[0105] The present application provides an electronic navigation map quality detection method and device to detect the quality problems of an electronic navigation map. The embodiments of the present application will be described in detail below.

[0106] Figure 1 A flowchart of an electronic navigation map quality detection method provided by an embodiment of the present application is shown in FIG. 1. The method can include the following steps:

[0107] S101: Obtain an electronic navigation map to be detected.

[0108] In the embodiments of the present application, the method can be applied to any type of electronic device with computing capability, which can be a server or a terminal device. The electronic device can be set in a vehicle as a vehicle-mounted device, or can not be set in a vehicle as a non-vehicle-mounted device, which is all acceptable.

[0109] In this step, the electronic device can obtain an electronic navigation map to be detected. The electronic navigation map can be any type of electronic map, and the electronic navigation map includes map semantic data. The map semantic data can include semantic information representing lane lines, lamp posts, and traffic signs included in a scene corresponding to the electronic navigation map. The semantic information can include information representing shapes, sizes, types, and positions of the lane lines, lamp posts, and traffic signs in the scene corresponding to the electronic navigation map.

[0110] In the embodiments of the present application, the semantic information representing the lane lines included in the scene corresponding to the electronic navigation map in the map semantic data can be referred to as a map lane line, the semantic information representing the lamp posts included in the scene corresponding to the electronic navigation map can be referred to as a map lamp post, and the semantic information representing the traffic signs included in the scene corresponding to the electronic navigation map can be referred to as a map traffic sign. The electronic navigation map further includes a map identifier corresponding to each map semantic data. The map identifier can be a serial number or a letter identifier, and has uniqueness in the electronic navigation map, so that a worker can accurately locate each map semantic data in the electronic navigation map.

[0111] S102: Obtain trajectory information of the vehicle during driving in the scene corresponding to the electronic navigation map.

[0112] The trajectory information includes a plurality of pose information and a positioning time corresponding to each pose information.

[0113] In the embodiments of the present application, the electronic navigation map quality detection process provided by the present application can be performed after the vehicle drives in the scene corresponding to the electronic navigation map. The trajectory information obtained by the electronic device can be trajectory information generated during the entire or partial driving process of the vehicle driving in the scene corresponding to the electronic navigation map. The trajectory information includes pose information of the vehicle collected at each time during the entire or partial driving process of the vehicle driving in the scene corresponding to the electronic navigation map.

[0114] The electronic navigation map quality detection process provided by the present application can also be performed during the driving of the vehicle in the scene corresponding to the electronic navigation map. In this case, the trajectory information can be trajectory information generated at a time before the current time during the driving of the vehicle in the scene corresponding to the electronic navigation map. That is, the trajectory information includes pose information of the vehicle generated at a time before the current time during the driving of the vehicle in the scene corresponding to the electronic navigation map.

[0115] In an implementation manner, the plurality of pieces of pose information can be measured by a pose determination device such as an inertial navigation system, a global positioning system, and / or a high-precision inertial navigation device arranged on the vehicle. The embodiments of the present application do not limit the obtaining manner of the plurality of pieces of pose information included in the trajectory information of the vehicle in the scene corresponding to the electronic navigation map during driving. Any obtaining manner of the plurality of pieces of pose information included in the trajectory information of the vehicle in the scene corresponding to the electronic navigation map during driving in the related art can be applied to the embodiments of the present application. The high-precision inertial navigation device includes a high-precision inertial navigation device, such as a fiber-optic gyroscope and an acceleration sensor.

[0116] S103: Obtain perception data determined by the vehicle during driving.

[0117] Each piece of perception data is data detected from a road image collected by an image collection device of the vehicle, and a collection time corresponding to each road image has a corresponding relationship with a positioning time corresponding to each piece of pose information.

[0118] In this step, during driving of the vehicle in the scene corresponding to the electronic navigation map, the image collection device arranged on the vehicle can photograph the environment during driving of the vehicle to collect road images. Then, a target detection model based on deep learning can be used to detect perception data in each road image, where the perception data can include information such as shape, type, and size of a target included in the road image. The target can include lane lines, streetlight poles, and traffic signboards. The target detection model can be a model trained based on sample images labeled with targets. The process of training the model can refer to the model training process in the related art, which will not be described here.

[0119] In the embodiments of the present application, the electronic device can directly obtain the perception data identified from the road images by other devices, or the electronic device can directly obtain the road images collected by the image collection device of the vehicle during driving of the vehicle in the scene corresponding to the electronic navigation map, and then detect the perception data in each road image based on the target detection model based on deep learning. This is also possible.

[0120] In the embodiments of the present application, the information representing lane lines in the image included in the perception data detected from the road images can be referred to as perceived lane lines, the information representing streetlight poles in the image included in the perception data detected from the road images can be referred to as perceived streetlight poles, and the information representing traffic signboards in the image included in the perception data detected from the road images can be referred to as perceived traffic signboards.

[0121] S104: For each piece of pose information, determine map semantic data corresponding to the pose information from the electronic navigation map.

[0122] It can be understood that each pose information of the vehicle in the process of driving in the scene corresponding to the electronic navigation map can represent the position and attitude of the vehicle in the scene corresponding to the electronic navigation map, and each pose information corresponds to a position and attitude in the electronic navigation map. The electronic device can determine the map area of the electronic navigation map corresponding to the pose information from the electronic navigation map through each pose information, and then determine the map semantic data corresponding to the pose information based on the map area corresponding to the pose information for each pose information.

[0123] S105: For each pose information, based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map, and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image, the quality detection result of the electronic navigation map is determined.

[0124] In this step, for each pose information, the corresponding perception data and the corresponding map semantic data in the electronic navigation map can be determined, wherein the perception data and the map semantic data have a corresponding relationship. Based on the observation position information of the perception data corresponding to the pose information in the road image, and / or the map position information of the map semantic data in the electronic navigation map, the detection result of the map semantic data in the electronic navigation map is determined, and the quality detection result of the electronic navigation map is determined based on the detection result of the map semantic data in the electronic navigation map. In order to layout clearly, the process of determining the quality detection result of the electronic navigation map is introduced in detail.

[0125] In the embodiment of the application, after the quality detection result of the electronic navigation map is determined, the quality detection result can be stored to facilitate subsequent staff to check the quality detection result of the electronic navigation map and correct the electronic navigation map based on the quality detection result.

[0126] By applying the embodiment of the application, the map semantic data and the perception data corresponding to each other in the electronic navigation map can be determined based on the positioning time corresponding to the pose information in the trajectory information of the vehicle in the process of driving in the scene corresponding to the electronic navigation map, and then the quality detection result of the electronic navigation map is determined based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map, and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image, to realize the detection of the quality problem of the electronic navigation map.

[0127] In another embodiment of the application, the quality detection result includes the detection result of the position deviation of each map semantic data; for example, Figure 2As shown, the S105 can include:

[0128] S201: For each pose information, based on the map semantic data corresponding to the pose information in the map position information of the electronic navigation map and the pose information, determine the projection position information of the map semantic data in the road image corresponding to the pose information.

[0129] S202: Based on the projection position information of the map semantic data in the road image corresponding to the pose information and the observation position information of the perception data corresponding to the map semantic data in the road image, determine the position deviation value between the map semantic data and the perception data corresponding thereto.

[0130] S203: Determine whether the position deviation value exceeds a preset distance threshold.

[0131] S204: If the result of the determination is that the position deviation value exceeds the preset distance threshold, determine that the map semantic data in the electronic navigation map has a position deviation.

[0132] In the embodiments of the present application, whether the position information of each map semantic data in the electronic navigation map is accurate can be detected, and based on the detection result of whether the position information of each map semantic data is accurate, the quality detection result of the electronic navigation map is determined.

[0133] The electronic device can determine, for each pose information, whether the map semantic data corresponding to the pose information in the map position information of the electronic navigation map is accurate, wherein for the convenience of description, the map position information of the map semantic data in the electronic navigation map can be referred to as the map position information, and the position information of the perception data in the corresponding road image can be referred to as the perception position information.

[0134] In an implementation, for each piece of pose information, the electronic device can determine, according to the piece of pose information and a previous piece of pose information of the piece of pose information, a pose change between the piece of pose information and the previous piece of pose information of the piece of pose information; obtain position information of perception data corresponding to the piece of pose information in a road image corresponding to the previous piece of pose information; and determine, by using a triangulation algorithm, depth information of the perception data based on the pose change between the piece of pose information and the previous piece of pose information of the piece of pose information and the position information of the perception data corresponding to the piece of pose information in the road image corresponding to the previous piece of pose information, i.e., determine depth information of map semantic data corresponding to the perception data. The depth information represents distance information between the perception data and an image collection device of the vehicle, and can also be regarded as distance information between the perception data and the vehicle, i.e., the depth information represents distance information between map semantic data corresponding to the perception data and the image collection device of the vehicle, and can also be regarded as distance information between the map semantic data corresponding to the perception data and the vehicle. Then, the electronic device calculates, based on the piece of pose information, map position information of the map semantic data corresponding to the piece of pose information in an electronic navigation map, and the depth information of the map semantic data, projection position information of the map semantic data in a road image corresponding to the piece of pose information.

[0135] Alternatively, the vehicle can be provided with a laser sensor, by which the depth information of the perception data, i.e., the depth information of map semantic data corresponding to the perception data, can be collected, the electronic device obtains the depth information of the map semantic data corresponding to the perception data collected by the laser sensor, and calculates, based on the piece of pose information, map position information of the map semantic data corresponding to the piece of pose information in an electronic navigation map, and the depth information of the map semantic data, projection position information of the map semantic data in a road image corresponding to the piece of pose information.

[0136] The process of calculating the projection position information of the map semantic data in the road image corresponding to the pose information based on the pose information, the map semantic data corresponding to the pose information, the map position information of the electronic navigation map, and the depth information of the map semantic data can be: converting the map semantic data corresponding to the pose information from a map coordinate system to a device coordinate system based on a first conversion relationship between the map coordinate system corresponding to the electronic navigation map and the device coordinate system of an image acquisition device corresponding to the road image corresponding to the pose information, i.e., determining the position information of the map semantic data corresponding to the pose information in the device coordinate system based on the first conversion relationship and the map position information of the map semantic data corresponding to the pose information in the electronic navigation map; and determining the projection position information of the map semantic data corresponding to the pose information in the road image corresponding to the pose information based on the position information of the map semantic data corresponding to the pose information in the device coordinate system, the depth information of the map semantic data, and a second conversion relationship between an image coordinate system corresponding to the road image corresponding to the pose information and the device coordinate system.

[0137] The second conversion relationship between the image coordinate system corresponding to the road image and the device coordinate system is a conversion relationship calibrated in advance for the image acquisition device, and the image acquisition device determines the second conversion relationship. The first conversion relationship can be determined based on the first conversion relationship between the map coordinate system corresponding to the electronic navigation map and the device coordinate system of the image acquisition device corresponding to the road image corresponding to the pose information.

[0138] In one case, the pose information is determined by a vehicle pose determination device, and the pose information is information describing the pose of the vehicle pose determination device, i.e., in the embodiment of the present application, the pose information of the vehicle pose determination device is used as the information describing the pose of the vehicle. To improve the accuracy of the quality detection result of the electronic navigation map to a certain extent, before S201, the pose information of the image acquisition device can be determined based on the conversion relationship between the coordinate system of the pose determination device and the device coordinate system of the image acquisition device, and the pose information of the pose determination device, and then for the pose information of each image acquisition device, the projection position information of the map semantic data in the road image corresponding to the pose information of the image acquisition device is determined based on the map position information of the map semantic data corresponding to the pose information of the image acquisition device in the electronic navigation map and the pose information of the image acquisition device; and the subsequent steps are performed.

[0139] Wherein, when the image acquisition device is arranged in the vehicle, the relative positional relationship between the image acquisition device and the vehicle and the relative positional relationship between the image acquisition device and the pose determination device arranged in the vehicle can be fixed, and the relative positional relationship between the image acquisition device and the vehicle and the relative positional relationship between the image acquisition device and the pose determination device arranged in the vehicle can be calibrated by the positional relationship calibration manner in the related art.

[0140] In one case, the map semantic data corresponding to each pose information can include at least one map semantic data, for example, can include at least one map lane line, at least one map street lamp pole and / or at least one map traffic sign, etc. The perception data corresponding to the pose information can include a plurality of at least one perception data, for example, can include at least one perception lane line, at least one perception street lamp pole and / or at least one perception traffic sign, etc. Wherein, for each pose information, each map semantic data and perception data have a one-to-one correspondence. Correspondingly, the electronic device can be for each map semantic data corresponding to each pose information, based on the projection position information of the map semantic data in the road image corresponding to the pose information, and the observation position information of the perception data corresponding to the map semantic data in the road image, calculate the corresponding distance between the map semantic data and the perception data corresponding thereto, which is the position deviation value between the map semantic data and the perception data corresponding thereto.

[0141] Further, for each map semantic data corresponding to each pose information, the position deviation value between the map semantic data and the perception target corresponding thereto is compared with the preset distance threshold to determine whether the position deviation value between the map semantic data and the perception data corresponding thereto exceeds the preset distance threshold. If it exceeds the preset distance threshold, it is determined that the projection position information of the map semantic data in the road image corresponding to the pose information and the observation position information of the perception data corresponding thereto in the road image are too far apart, and further, it is determined that the position of the map semantic data in the electronic navigation map has a position deviation.

[0142] In one case, when the position of at least one map semantic data in the electronic navigation map is determined to have a position deviation for each map semantic data corresponding to each pose information, it can be further determined that the electronic navigation map has a quality problem. Wherein, the quality detection result corresponding to the electronic navigation map includes information representing that the map semantic data has a position deviation, wherein it can be understood that in order to ensure that the staff can quickly locate which map lane lines in the electronic navigation map have problems, the map identifier corresponding to the map semantic data having the above position deviation and the map position information of the map semantic data in the electronic navigation map can be recorded accordingly.

[0143] In another implementation of the present application, if the determination result is that the position deviation values corresponding to all the map semantic data in the electronic navigation map do not exceed the preset distance threshold, it can be determined that there is no position deviation of the map semantic data in the electronic navigation map.

[0144] In another embodiment of the present application, the perception data includes a perception lane line, and the map semantic data includes a map lane line; the quality detection result includes a detection result of position deviation of the map lane line in the elevation direction; and the S105 can include: for each pose information, determining, based on the map position information of the map lane line in the electronic navigation map in the map semantic data corresponding to the pose information and the pose information, mapping position information of the map lane line in a vehicle coordinate system corresponding to the pose information.

[0145] For each pose information, based on the observation position information of the perception lane line in the corresponding road image in the perception data corresponding to the pose information, the pose information and the projection matrix corresponding to the image acquisition device, the mapping position information of the perception lane line in the vehicle coordinate system corresponding to the pose information is determined; wherein the projection matrix is the second conversion relationship described above.

[0146] For each map lane line corresponding to each pose information, based on the mapping position information of the map lane line in the vehicle coordinate system corresponding to the pose information and the mapping position information of the perception lane line corresponding to the map lane line in the vehicle coordinate system corresponding to the pose information, it is determined whether there is a position deviation of the map lane line in the horizontal axis direction and the vertical axis direction of the vehicle coordinate system.

[0147] If it is determined that there is no position deviation of the map lane line in the horizontal axis direction and the vertical axis direction of the vehicle coordinate system, for each pose information, based on the map position information of the map lane line in the electronic navigation map in the map semantic data corresponding to the pose information and the pose information, the projection position information of the map lane line in the road image corresponding to the pose information is determined.

[0148] For each map lane line corresponding to each pose information, based on the projection position information of the map lane line in the road image corresponding to the pose information and the observation position information of the perception lane line in the road image in the perception data corresponding to the map lane line, it is determined whether there is a position deviation of the map lane line in the elevation direction.

[0149] The vehicle body coordinate system is also called a wheel speed coordinate system, which can take the midpoint of the center line of the two rear wheels of the vehicle as the origin, and can be observed from the tail to the head, from left to right as the direction of the horizontal axis of the vehicle body coordinate system, from back to front as the direction of the vertical axis of the vehicle body coordinate system, and from bottom to top as the direction of the vertical axis of the vehicle body coordinate system. The elevation direction can be referred to as the vertical axis direction of the vehicle body coordinate system, the horizontal axis direction of the vehicle body coordinate system can be referred to as the left-right direction of the vehicle, and the vertical axis direction of the vehicle body coordinate system can be referred to as the front-rear direction of the vehicle.

[0150] In the present embodiment, when the perception data includes perceived lane lines and the map semantic data includes map lane lines, the position deviation of the map lane lines in the electronic navigation map can be more finely determined, and the direction of the position deviation can be determined, so as to obtain a more finely determined quality detection result, and facilitate subsequent correction of the electronic navigation map by the staff.

[0151] The electronic device determines, for each pose information, a vehicle body coordinate system corresponding to the pose information based on the pose information; further determines a conversion relationship between the vehicle body coordinate system corresponding to the pose information and the electronic navigation map as a third conversion relationship; and determines, based on the third conversion relationship and map position information of a map lane line in the map semantic data corresponding to the pose information in the electronic navigation map, mapping position information of the map lane line in the map semantic data corresponding to the pose information in the vehicle body coordinate system corresponding to the pose information.

[0152] For each pose information, the electronic device determines the pose change between the pose information and the previous pose information of the pose information according to the pose information and the previous pose information of the pose information; obtains position information of a perceived lane line corresponding to the pose information in a road image corresponding to the previous pose information of the pose information; and determines, by using a triangulation algorithm, depth information of the perceived lane line based on the pose change between the pose information and the previous pose information of the pose information and the position information of the perceived lane line corresponding to the pose information in the road image corresponding to the previous pose information of the pose information. Alternatively, the vehicle can be provided with a laser sensor, and the depth information of the perceived lane line can be collected by the laser sensor. The electronic device obtains the depth information of the perceived lane line collected by the laser sensor.

[0153] The depth information represents the distance information between the perceived lane line and the image collection device of the vehicle, and can also be regarded as the distance information between the perceived lane line and the vehicle.

[0154] The electronic device calculates, based on the pose information, observation position information of the perception lane line corresponding to the pose information in the corresponding road image, and depth information of the perception data, device position information of the perception lane line in a device coordinate system corresponding to the road image corresponding to the pose information. Based on a conversion relationship between the device coordinate system corresponding to the road image corresponding to the pose information and the vehicle body coordinate system, and the device position information of the perception lane line in the device coordinate system corresponding to the road image corresponding to the pose information, mapping position information of the perception lane line in the vehicle body coordinate system corresponding to the pose information is determined.

[0155] In an implementation manner, each map lane line in each map semantic data in the electronic navigation map can include a series of discrete points, and the mapping position information of each map lane line in the vehicle body coordinate system corresponding to the pose information includes mapping position information of a plurality of discrete points corresponding to the map lane line.

[0156] Correspondingly, the electronic device can calculate, for each discrete point included in each map lane line corresponding to each pose information, an error sub-vector between the discrete point and the perception lane line corresponding to the map lane line based on the mapping position information of the discrete point and the mapping position information of the perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information, wherein a size of a module of the error sub-vector between the discrete point and the perception lane line corresponding to the map lane line is equal to a distance between the discrete point and the perception lane line corresponding to the map lane line, and a direction of the error sub-vector between the discrete point and the perception lane line corresponding to the map lane line points to the discrete point.

[0157] Subsequently, the electronic device determines, based on the error sub-vector between each discrete point included in each map lane line corresponding to each pose information and the perception lane line corresponding to the map lane line, an error vector between each map lane line corresponding to each pose information and the perception lane line corresponding to the map lane line as an error vector corresponding to each map lane line.

[0158] Wherein, a size of a module of the error vector between each map lane line corresponding to each pose information and the perception lane line corresponding to the map lane line is equal to an average value of a module of each error sub-vector between each discrete point included in the map lane line and the perception lane line corresponding to the map lane line, and a direction of the error vector between each map lane line corresponding to each pose information and the perception lane line corresponding to the map lane line is a direction of any specified error sub-vector in the error sub-vectors between each discrete point included in the map lane line and the perception lane line corresponding to the map lane line. Wherein, the specified error sub-vector can be an error sub-vector corresponding to a discrete point located at a middle position included in the map lane line.

[0159] The electronic device determines, for each map lane line corresponding to each piece of pose information, an error vector component of an error vector corresponding to the map lane line in a lateral axis direction of a vehicle coordinate system corresponding to the piece of pose information and an error vector component of the error vector in a longitudinal axis direction of the vehicle coordinate system, based on the error vector corresponding to the map lane line, and calculates a module of the error vector component of the error vector in the lateral axis direction of the vehicle coordinate system and a module of the error vector component of the error vector in the longitudinal axis direction of the vehicle coordinate system, respectively.

[0160] The electronic device determines whether the module of the error vector component of the error vector corresponding to the map lane line in the lateral axis direction of the vehicle coordinate system exceeds a first value, and if the module of the error vector component of the error vector corresponding to the map lane line in the lateral axis direction of the vehicle coordinate system does not exceed the first value, determines that there is no position deviation of the map lane line in the lateral axis direction of the vehicle coordinate system, and otherwise, if the module of the error vector component of the error vector corresponding to the map lane line in the lateral axis direction of the vehicle coordinate system exceeds the first value, determines that there is a position deviation of the map lane line in the lateral axis direction of the vehicle coordinate system.

[0161] The electronic device determines whether the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle coordinate system exceeds a second value, and if the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle coordinate system does not exceed the second value, determines that there is no position deviation of the map lane line in the longitudinal axis direction of the vehicle coordinate system, and otherwise, if the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle coordinate system exceeds the first value, determines that there is a position deviation of the map lane line in the longitudinal axis direction of the vehicle coordinate system.

[0162] The first value and the second value can be values set according to actual conditions, and the first value and the second value can be equal or not equal.

[0163] In a case where the electronic device determines that the map lane line does not exist position deviation in the horizontal axis and the vertical axis of the vehicle body coordinate system, the electronic device can continue to project the map lane line in the map semantic data corresponding to each pose information into the road image corresponding to the pose information, i.e., determine the projection position information of the map lane line in the road image corresponding to the pose information based on the map position information of the map lane line in the map semantic data corresponding to the pose information and the pose information, and then determine whether the map lane line exists position deviation in the elevation direction based on the projection position information of the map lane line in the road image corresponding to the pose information and the observation position information of the perception lane line in the road image corresponding to the perception data of the map lane line.

[0164] In an implementation manner, after the electronic device determines that the map lane line exists position deviation in the horizontal axis and / or the vertical axis of the vehicle body coordinate system, the electronic device can no longer perform the subsequent step of determining whether the map lane line exists position deviation in the elevation direction of the vehicle body coordinate system.

[0165] In a case, the mapping position information of the perception lane line corresponding to each map lane line in the vehicle body coordinate system corresponding to the pose information can include the mapping position information of a plurality of discrete points corresponding to the perception lane line. Correspondingly, the electronic device can calculate the error sub-vector between each discrete point corresponding to each map lane line and the perception lane line corresponding to the map lane line based on the mapping position information of the discrete point and the mapping position information of the perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information.

[0166] The electronic device can traverse the projection points of the plurality of discrete points included in the perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information, determine the two projection points closest to the discrete point from the projection points of the plurality of discrete points included in the perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information, calculate the error sub-vector between the discrete point and the line connecting the two projection points closest to the discrete point, and take the error sub-vector between the discrete point and the line connecting the two projection points closest to the discrete point as the error sub-vector between the discrete point and the perception lane line corresponding to the map lane line.

[0167] Further, the quality detection result corresponding to the electronic navigation map includes information representing a position deviation of the map lane line in a height direction, a position deviation of the vehicle in a left-right direction, or a position deviation of the vehicle in a front-rear direction, wherein it can be understood that, in order to enable the staff to quickly locate which map lane line in the electronic navigation map has a problem, the map identifier corresponding to the map lane line having the above-mentioned problem and the map position information of the map lane line in the electronic navigation map can be recorded correspondingly.

[0168] In another embodiment of the present application, the perception data includes a perceived lane line, and the map semantic data includes a map lane line; the quality detection result includes a detection result of a bending condition and / or a broken condition of the map lane line.

[0169] In an implementation manner, if the quality detection result includes the detection result of the bending condition of the map lane line, the S105 can include:

[0170] For each pose information, based on the map position information of the map lane line in the electronic navigation map in the map semantic data corresponding to the pose information and the pose information, mapping position information of the map lane line in a vehicle coordinate system corresponding to the pose information is determined, wherein each mapping position information includes mapping position information of a plurality of discrete points corresponding to the corresponding map lane line.

[0171] For each map lane line corresponding to each pose information, based on the mapping position information of the plurality of discrete points corresponding to the map lane line and a preset fitting algorithm, a lane line fitting line corresponding to the map lane line is fitted;

[0172] For each map lane line corresponding to each pose information, based on the lane line fitting line corresponding to the map lane line and the mapping position information of the plurality of discrete points corresponding to the map lane line, distance variance corresponding to the map lane line is determined.

[0173] If the distance variance exceeds a preset variance threshold, it is determined that the map lane line has a lane line bending condition.

[0174] In the embodiment of the present application, the bending condition of the lane line in the electronic navigation map can also be detected, and the quality detection result includes a detection result of the lane line bending condition.

[0175] In an implementation manner, each map lane line in each map semantic data in the electronic navigation map can include a series of discrete points. In theory, a series of discrete points included in each map lane line are uniformly distributed and distributed along a straight line. In the embodiment of the present application, the electronic device can map, for each map lane line corresponding to each pose information, a series of discrete points included in the map lane line into a vehicle coordinate system corresponding to the pose information, to obtain mapping position information of the series of discrete points included in the map lane line in the vehicle coordinate system corresponding to the pose information, and based on the mapping position information of the series of discrete points included in each map lane line in the vehicle coordinate system corresponding to the pose information and a preset fitting algorithm, a lane line fitting line corresponding to the map lane line in the vehicle coordinate system corresponding to the pose information is fitted. The preset fitting algorithm can be any type of straight line fitting method such as least square method straight line fitting method.

[0176] In theory, if each map lane line does not appear a bending condition, the distance variance between the mapping position information of the series of discrete points included in each map lane line in the vehicle coordinate system corresponding to the pose information and the lane line fitting line corresponding to the map lane line in the vehicle coordinate system corresponding to the pose information is not large.

[0177] Therefore, for each map lane line corresponding to each pose information, the electronic device calculates the square of the distance between each discrete point corresponding to the map lane line and the lane line fitting line corresponding to the map lane line based on the mapping position information of the series of discrete points corresponding to the map lane line and the lane line fitting line corresponding to the map lane line, and then calculates the sum of the squares of the distances between each discrete point corresponding to the map lane line and the lane line fitting line corresponding to the map lane line as the distance variance corresponding to the map lane line, to determine whether the distance variance exceeds a preset variance threshold. If the distance variance exceeds the preset variance threshold, it is determined that the map lane line has a lane line bending condition. If the distance variance does not exceed the preset variance threshold, it is determined that the map lane line does not have a lane line bending condition. The preset variance threshold is a value set according to the situation.

[0178] In an implementation manner, if the quality detection result includes a detection result of a map lane line breakage condition, the S105 can include: determining, for each pose information, conversion position information of a map lane line in a vehicle coordinate system corresponding to the pose information or a road image based on map position information of the map lane line in the electronic navigation map and the pose information corresponding to the map position information, wherein each conversion position information includes conversion position information of a plurality of discrete points corresponding to the corresponding map lane line.

[0179] For each map lane line corresponding to each pose information, based on the conversion position information of the plurality of discrete points corresponding to the map lane line, the distance between each adjacent two discrete points is calculated.

[0180] If the distance between the adjacent two discrete points in the plurality of discrete points corresponding to the map lane line exceeds the preset difference value from the distance between the other adjacent two discrete points, it is determined that the map lane line has a lane line break condition.

[0181] In an implementation manner, each map lane line in each map semantic data in the electronic navigation map can include a series of discrete points. In theory, the series of discrete points included in each map lane line are uniformly distributed and distributed along a straight line. In a case, the electronic device can map the series of discrete points included in each map lane line corresponding to each pose information to the vehicle coordinate system corresponding to the pose information, to obtain the mapping position information of the series of discrete points included in the map lane line in the vehicle coordinate system corresponding to the pose information. Alternatively, the electronic device can project the series of discrete points included in each map lane line corresponding to each pose information into the road image corresponding to the pose information, to obtain the projection position information of the series of discrete points included in the map lane line in the road image corresponding to the pose information. In the embodiment of the application, the mapping position information of the series of discrete points included in the map lane line in the vehicle coordinate system corresponding to the pose information and the projection position information of the series of discrete points included in the map lane line in the road image corresponding to the pose information can be referred to as conversion position information.

[0182] Further, the electronic device calculates the distance between each adjacent two discrete points based on the conversion position information of the plurality of discrete points corresponding to each map lane line corresponding to each pose information, and compares the size of the distance between each adjacent two discrete points. In theory, the series of discrete points included in each map lane line in the electronic navigation map are uniformly distributed, and after being mapped into the road image corresponding to the corresponding pose information or the vehicle coordinate system, the distance between the conversion position information of each discrete point corresponding to the map lane line is also uniformly distributed. Therefore, if the distance between the adjacent two discrete points in the plurality of discrete points corresponding to the map lane line exceeds the preset difference value from the distance between the other adjacent two discrete points, it can be determined that the series of discrete points included in the map lane line in the electronic navigation map are not uniformly distributed, and further, it is determined that the map lane line has a lane line break condition.

[0183] Further, the quality detection result corresponding to the electronic navigation map includes information representing that the lane line in the map lane has a lane line breakage condition. It can be understood that, in order to enable the staff to quickly locate which map lane line in the electronic navigation map has a problem, the map identifier corresponding to the map lane line having the lane line breakage condition and the map position information of the map lane line in the electronic navigation map can be recorded accordingly.

[0184] In another embodiment of the present application, the perception data includes a perception street lamp pole, and the map semantic data includes a map street lamp pole; the quality detection result includes a detection result of a tilt condition of the map street lamp pole; and the S105 can include:

[0185] For each pose information, based on the map position information of the map street lamp pole in the map semantic data corresponding to the pose information in the electronic navigation map and the pose information, projection position information of the map street lamp pole in a road image corresponding to the pose information is determined.

[0186] For each map street lamp pole corresponding to each pose information, based on the projection position information of the map street lamp pole in a road image corresponding to the pose information and the observation position information of the perception street lamp pole corresponding to the map street lamp pole in the corresponding road image, it is determined whether the map street lamp pole has a tilt condition.

[0187] In the present implementation, if the perception data includes a perception street lamp pole and the map semantic data includes a map street lamp pole, the electronic device needs to detect whether the map street lamp pole has a tilt condition. During the detection process, the electronic device can project the map street lamp pole in the map semantic data corresponding to each pose information into a road image corresponding to the pose information, i.e., based on the map position information of the map street lamp pole in the map semantic data corresponding to the pose information in the electronic navigation map and the pose information, projection position information of the map street lamp pole in a road image corresponding to the pose information is determined; and the observation position information of the perception street lamp pole corresponding to each map street lamp pole in the corresponding road image, i.e., the road image corresponding to the pose information, is determined. For each map street lamp pole corresponding to each pose information, based on the projection position information of the map street lamp pole in the road image corresponding to the pose information and the observation position information of the perception street lamp pole corresponding to the map street lamp pole in the road image corresponding to the pose information, an included angle between the map street lamp pole and the perception street lamp pole is calculated, and whether the map street lamp pole has a tilt condition is determined based on the included angle.

[0188] In one case, the electronic device can determine, for each map road lamp pole corresponding to each pose information, whether an included angle between the map road lamp pole and the perceived road lamp pole exceeds a preset included angle, and if the included angle exceeds the preset included angle, it is determined that the map road lamp pole has an inclination condition, and if the included angle does not exceed the preset included angle, it is determined that the map road lamp pole does not have an inclination condition.

[0189] Further, if it is determined that the map road lamp pole has an inclination condition, the quality detection result corresponding to the electronic navigation map includes information representing that the map road lamp pole has an inclination condition. It can be understood that, in order to enable the staff to quickly locate which map road lamp poles in the electronic navigation map have problems, the map identifier corresponding to the map road lamp pole having an inclination condition and the map position information of the map road lamp pole in the electronic navigation map can be recorded accordingly.

[0190] In another embodiment of the present application, the perception data includes a perceived traffic sign, and the map semantic data includes a map traffic sign; the quality detection result includes a detection result of a map traffic sign overturning condition;

[0191] The S105 can include: for each pose information, determining, based on map position information of a map traffic sign in the map semantic data corresponding to the pose information in the electronic navigation map and the pose information, mapping position information of the map traffic sign in a vehicle coordinate system corresponding to the pose information;

[0192] For each pose information, based on observation position information of a perceived traffic sign in the perception data corresponding to the pose information in a corresponding road image, the pose information, and a projection matrix corresponding to an image acquisition device, mapping position information of the perceived traffic sign in a vehicle coordinate system corresponding to the pose information is determined.

[0193] For each map traffic sign corresponding to each pose information, based on mapping position information of the map traffic sign in a vehicle coordinate system corresponding to the pose information, and mapping position information of a perceived traffic sign corresponding to the map traffic sign in a vehicle coordinate system corresponding to the pose information, it is determined whether the map traffic sign has an overturning condition.

[0194] In the present implementation, if the perception data includes a perceived traffic sign and the map semantic data includes a map traffic sign, the electronic device needs to detect whether the map traffic sign has an overturning condition. In the detection process, the electronic device maps the map traffic sign in the electronic navigation map and the perceived map traffic sign corresponding to the map traffic sign into a vehicle coordinate system to determine whether the map traffic sign has an overturning condition through the spatial coordinate system.

[0195] Specifically, the electronic device maps the map traffic sign in the map position information of the electronic navigation map to the vehicle body coordinate system corresponding to the pose information based on the map position information of the map traffic sign in the map semantic data corresponding to the pose information and the pose information, to determine the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information; and maps the perception traffic sign corresponding to the map traffic sign to the vehicle body coordinate system corresponding to the pose information based on the observation position information of the perception traffic sign in the perception data corresponding to the pose information, the pose information and the projection matrix corresponding to the image acquisition device, to determine the mapping position information of the perception traffic sign in the vehicle body coordinate system corresponding to the pose information; wherein the projection matrix corresponding to the image acquisition device is the second conversion relationship described above. Further, based on the map traffic sign corresponding to each pose information, based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information, and the mapping position information of the perception traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information, it is determined whether the map traffic sign is in a flipped state.

[0196] The process of mapping the map traffic sign to the vehicle body coordinate system corresponding to the pose information can refer to the process of mapping the map lane line to the vehicle body coordinate system corresponding to the pose information described above, and the process of mapping the perception traffic sign corresponding to the map traffic sign to the vehicle body coordinate system corresponding to the pose information can refer to the process of mapping the perception lane line corresponding to the map lane line to the vehicle body coordinate system corresponding to the pose information described above, which will not be described again here.

[0197] In an embodiment of the present application, the step of determining whether the map traffic sign is in a flipped state based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information can include:

[0198] fitting a first fitting surface corresponding to the map traffic sign based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information and a preset plane fitting algorithm;

[0199] fitting a second fitting surface corresponding to the perception traffic sign corresponding to the map traffic sign based on the mapping position information of the perception traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information and the preset plane fitting algorithm;

[0200] calculating the included angle between the normal vector corresponding to the first fitting surface and the normal vector corresponding to the second fitting surface;

[0201] If the included angle exceeds the preset angle, it is determined that the map traffic sign exists a flip situation.

[0202] In the present implementation, the electronic device can fit a first fitting plane corresponding to the map traffic sign based on the mapping position information of the map traffic sign in the vehicle coordinate system corresponding to the pose information and a preset plane fitting algorithm, and fit a second fitting plane corresponding to the perceived traffic sign corresponding to the map traffic sign based on the mapping position information of the perceived traffic sign corresponding to the map traffic sign in the vehicle coordinate system corresponding to the pose information and the preset plane fitting algorithm. Further, the normal vector of the first fitting plane and the normal vector of the second fitting plane are determined, and the included angle between the normal vector of the first fitting plane and the normal vector of the second fitting plane is calculated. The included angle between the normal vector of the first fitting plane and the normal vector of the second fitting plane is compared with the preset angle. If the included angle between the normal vector of the first fitting plane and the normal vector of the second fitting plane exceeds the preset angle, it is determined that the map traffic sign exists a flip situation. If the included angle between the normal vector of the first fitting plane and the normal vector of the second fitting plane does not exceed the preset angle, it is determined that the map traffic sign does not exist a flip situation.

[0203] In one case, the traffic sign is generally a plane, and the preset plane fitting algorithm can be any type of algorithm that can fit a plane based on a series of spatial points, such as a plane fitting algorithm based on least squares.

[0204] In the case where it is determined that the map traffic sign exists a flip situation, the quality detection result corresponding to the electronic navigation map includes information representing that the map traffic sign exists a flip situation. It can be understood that, in order to ensure that the staff can quickly locate which map traffic signs in the electronic navigation map have problems, the map identifier corresponding to the map traffic sign that exists a flip situation and the map position information of the map traffic sign in the electronic navigation map can be recorded accordingly.

[0205] In another embodiment of the present application, after S105, the method can further include:

[0206] If the quality detection result of the electronic navigation map represents that the electronic navigation map exists a quality problem, the abnormal statistical document and the abnormal image information corresponding to the electronic navigation map are stored, wherein the abnormal statistical document at least includes: the identifier of the map semantic data in the electronic navigation map that exists a quality problem and the problem quality type thereof, and the map position information in the electronic navigation map, and the abnormal image information includes: the image information corresponding to the map semantic data that exists a quality problem.

[0207] If the quality detection result of the electronic navigation map indicates that the electronic navigation map has a quality problem, that is, at least one map semantic data in the electronic navigation map has a quality problem, the quality problem of the map semantic data can be classified. The problem quality type obtained by classifying the quality problem of the map semantic data can include, but is not limited to, position deviation, map lane line bending condition, breakage condition, map street lamp pole tilting condition, and map traffic sign plate overturning condition. Specifically, the classification can be further refined. For example, when the map lane line has a position deviation, the problem quality type corresponding to the map lane line is a position deviation, and the position deviation can be further classified into a position deviation in the elevation direction, a position deviation in the left-right direction of the vehicle, and a position deviation in the front-rear direction of the vehicle.

[0208] In one case, for each map semantic data in the electronic navigation map, the image acquisition device of the vehicle can observe the real target corresponding to the map semantic data multiple times during the driving of the vehicle in the scene corresponding to the electronic navigation map, that is, each map semantic data can correspond to multiple pose information. Accordingly, for different pose information, the electronic device can determine a position deviation, a position deviation in the elevation direction, a map lane line bending condition, a map lane line breakage condition, a map street lamp pole tilting condition, and / or a map traffic sign plate overturning condition detection result of the map semantic data. The position deviation, the position deviation in the elevation direction, the map lane line bending condition, the map lane line breakage condition, the map street lamp pole tilting condition, and / or the map traffic sign plate overturning condition can be referred to as an error condition of the map semantic data.

[0209] At this time, for each map semantic data, the error condition with the largest error of the map semantic data can be determined from each problem quality type and the error condition corresponding to the map semantic data, and the largest error condition corresponding to each map semantic data can be added to the exception statistical document.

[0210] The error case representing the largest position deviation of the map semantic data can be an error case with the largest error value. For example, when the error case is a position deviation case, the error value can be a position deviation value; when the error case is a position deviation case in the height direction, the error value is a position deviation value in the height direction; when the error case is a map lane line bending condition, the error value is a distance variance of the map lane line; when the error case is a map lane line breakage condition, the error value is a difference between a distance between two adjacent discrete points of the map lane line and a distance between other two adjacent discrete points; when the error case is a map street lamp pole inclination condition, the error value is an included angle between the map street lamp pole and the corresponding perceived street lamp pole; and when the error case is a map traffic sign plate overturning condition, the error value is an included angle between the map traffic sign plate and the corresponding perceived traffic sign plate.

[0211] The abnormal image information can be, for example, a road image containing a projection point of map semantic data with a quality problem. The abnormal image information reflects a comparison result between map semantic data and perception data with a corresponding relationship, and can be used for subsequent analysis of specific quality problems of map semantic data with a quality problem in the electronic navigation map.

[0212] In an implementation manner, the position deviation case of the map lane line in the map semantic data in the electronic navigation map can be detected first, then the lane line bending condition and the lane line breakage condition of the map lane line in the map semantic data are detected, then the map street lamp pole inclination condition of the map street lamp pole in the map semantic data is detected, then the map traffic sign plate overturning condition of the map traffic sign plate in the map semantic data is detected, and finally the position precision of the map semantic data is detected, that is, the position deviation case of the map semantic data is detected. It can be understood that the above is only an example of the execution order of the detection aspect, and does not limit the execution order of the above detection aspect provided by the embodiments of the present application.

[0213] Corresponding to the method embodiments, the embodiments of the present application provide an electronic navigation map quality detection device, as shown in Figure 3 may include:

[0214] The first obtaining module 310 is configured to obtain an electronic navigation map to be detected.

[0215] The second obtaining module 320 is configured to obtain trajectory information of a vehicle during driving in a scene corresponding to the electronic navigation map, wherein the trajectory information includes a plurality of pose information and a positioning time corresponding to each pose information.

[0216] The third obtaining module 330 is configured to obtain perception data determined by the vehicle during the driving process, wherein each piece of perception data is data detected from a road image collected by an image collection device of the vehicle, and each road image corresponds to a collection time instant corresponding to each piece of pose information;

[0217] The first determining module 340 is configured to determine, for each piece of pose information, map semantic data corresponding to the pose information from the electronic navigation map;

[0218] The second determining module 350 is configured to determine, for each piece of pose information, a quality detection result of the electronic navigation map based on map location information of the map semantic data corresponding to the pose information in the electronic navigation map and / or observation location information of the perception data corresponding to the pose information in the corresponding road image.

[0219] According to the embodiments of the present application, the map semantic data and the perception data corresponding to each other in the electronic navigation map can be determined based on the pose information corresponding to the positioning time instant in the trajectory information of the vehicle in the scene corresponding to the electronic navigation map during the driving process, and then, for each piece of pose information, the quality detection result of the electronic navigation map can be determined based on the map location information of the map semantic data corresponding to the pose information in the electronic navigation map and / or the observation location information of the perception data corresponding to the pose information in the corresponding road image, so as to realize the detection of the quality problem of the electronic navigation map.

[0220] In another embodiment of the present application, the perception data includes a perception lane line, and the map semantic data includes a map lane line; and the quality detection result includes a detection result of a position deviation of the map lane line in the elevation direction.

[0221] The second determining module 350 is specifically configured to determine, for each piece of pose information, mapping location information of the map lane line in a vehicle coordinate system corresponding to the pose information based on the map location information of the map lane line in the electronic navigation map and the pose information in the map semantic data corresponding to the pose information.

[0222] For each piece of pose information, the mapping location information of the perception lane line in the vehicle coordinate system corresponding to the pose information is determined based on the observation location information of the perception lane line in the corresponding road image in the perception data corresponding to the pose information, the pose information, and a projection matrix corresponding to the image collection device.

[0223] For each map lane line corresponding to each pose information, based on the mapping position information of the map lane line in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information, it is determined whether there is a position deviation of the map lane line in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system;

[0224] If it is determined that there is no position deviation of the map lane line in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system, for each pose information, based on the map position information of the map lane line in the electronic navigation map in the map semantic data corresponding to the pose information and the pose information, the projection position information of the map lane line in the road image corresponding to the pose information is determined.

[0225] For each map lane line corresponding to each pose information, based on the projection position information of the map lane line in the road image corresponding to the pose information and the observation position information of the perception lane line in the road image in the perception data corresponding to the map lane line, it is determined whether there is a position deviation of the map lane line in the elevation direction.

[0226] In another embodiment of the present application, the perception data includes perception lane lines, and the map semantic data includes map lane lines; the quality detection result includes the detection result of the bending condition and / or the broken condition of the map lane line;

[0227] If the quality detection result includes the detection result of the bending condition of the map lane line, the second determination module 350 is specifically configured to, for each pose information, based on the map position information of the map lane line in the electronic navigation map in the map semantic data corresponding to the pose information and the pose information, determine the mapping position information of the map lane line in the vehicle body coordinate system corresponding to the pose information, wherein each mapping position information includes the mapping position information of a plurality of discrete points corresponding to the corresponding map lane line;

[0228] For each map lane line corresponding to each pose information, based on the mapping position information of a plurality of discrete points corresponding to the map lane line and a preset fitting algorithm, a lane line fitting line corresponding to the map lane line is fitted;

[0229] For each map lane line corresponding to each pose information, based on the lane line fitting line corresponding to the map lane line and the mapping position information of a plurality of discrete points corresponding to the map lane line, a distance variance corresponding to the map lane line is determined.

[0230] If the distance variance exceeds a preset variance threshold, it is determined that the map lane line has a lane line bending condition;

[0231] And / or, if the quality detection result includes a detection result of a map lane line breakage case, the second determination module 350 is specifically configured to, for each pose information, determine, based on map lane line in map semantic data corresponding to the pose information at map position information of the electronic navigation map and the pose information, conversion position information of the map lane line in a vehicle body coordinate system corresponding to the pose information or in a road image, wherein each conversion position information includes conversion position information of a plurality of discrete points corresponding to the corresponding map lane line.

[0232] For each map lane line corresponding to each pose information, based on conversion position information of a plurality of discrete points corresponding to the map lane line, the distance between each adjacent two discrete points is calculated.

[0233] If the distance between the adjacent two discrete points among the plurality of discrete points corresponding to the map lane line exceeds the preset difference value from the distance between other adjacent two discrete points, it is determined that the map lane line has a lane line breakage case.

[0234] In another embodiment of the application, the perception data includes a perception street lamp pole, and the map semantic data includes a map street lamp pole; the quality detection result includes a detection result of a map street lamp pole tilt condition;

[0235] The second determination module 350 is specifically configured to, for each pose information, determine, based on map street lamp pole in map semantic data corresponding to the pose information at map position information of the electronic navigation map and the pose information, projection position information of the map street lamp pole in a road image corresponding to the pose information;

[0236] For each map street lamp pole corresponding to each pose information, based on projection position information of the map street lamp pole in a road image corresponding to the pose information and observation position information of the perception street lamp pole in the corresponding road image in the perception data corresponding to the map street lamp pole, it is determined whether the map street lamp pole has a tilt condition.

[0237] In another embodiment of the application, the perception data includes a perception traffic sign, and the map semantic data includes a map traffic sign; the quality detection result includes a detection result of a map traffic sign flip case;

[0238] The second determination module 350 is specifically configured to, for each pose information, determine, based on map traffic sign in map semantic data corresponding to the pose information at map position information of the electronic navigation map and the pose information, mapping position information of the map traffic sign in a vehicle body coordinate system corresponding to the pose information;

[0239] For each pose information, based on observation position information of a perceived traffic sign in a corresponding road image in perception data corresponding to the pose information, the pose information, and a projection matrix corresponding to the image collection device, mapping position information of the perceived traffic sign in a vehicle body coordinate system corresponding to the pose information is determined.

[0240] For a map traffic sign corresponding to each pose information, based on mapping position information of the map traffic sign in a vehicle body coordinate system corresponding to the pose information, and mapping position information of a perceived traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information, whether the map traffic sign has a flip situation is determined.

[0241] In another embodiment of the present application, the second determination module 350 is specifically configured to fit a first fitting surface corresponding to the map traffic sign based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information and a preset plane fitting algorithm.

[0242] fit a second fitting surface corresponding to the perceived traffic sign corresponding to the map traffic sign based on the mapping position information of the perceived traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information and the preset plane fitting algorithm.

[0243] Calculate the included angle between the normal vector corresponding to the first fitting surface and the normal vector corresponding to the second fitting surface.

[0244] If the included angle exceeds a preset angle, it is determined that the map traffic sign has a flip situation.

[0245] In another embodiment of the present application, the device further comprises:

[0246] The storage module is configured to, after the quality detection result of the electronic navigation map is determined based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map, and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image, if the quality detection result of the electronic navigation map indicates that the electronic navigation map has a quality problem, store an abnormal statistical document and abnormal image information corresponding to the electronic navigation map, wherein the abnormal statistical document at least includes an identifier of the map semantic data having a quality problem in the electronic navigation map and a problem quality type thereof, map position information in the electronic navigation map, and the abnormal image information includes image information corresponding to the map semantic data having a quality problem.

[0247] The device and system embodiments correspond to the method embodiments and have the same technical effects as the method embodiments. For details, refer to the method embodiments. The device embodiments are based on the method embodiments. For details, refer to the method embodiments, which will not be described herein again.

[0248] Those skilled in the art can understand that the drawings are only schematic diagrams of an embodiment, and the modules or flows in the drawings are not necessarily required for implementing the present application.

[0249] Those skilled in the art can understand that the modules in the device in the embodiments can be distributed in the device in the embodiments as described in the embodiments, or can be correspondingly changed and located in one or more devices different from the embodiments. The modules in the above embodiments can be combined into one module, or can be further split into multiple sub-modules.

[0250] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can still be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An electronic navigation map quality detection method, characterized by, The method comprises: obtaining an electronic navigation map to be detected; obtaining trajectory information of a vehicle in a driving process of a scene corresponding to the electronic navigation map, wherein the trajectory information comprises a plurality of pose information and a positioning time corresponding to each pose information; obtaining perception data determined by the vehicle in the driving process, wherein each perception data is data detected from a road image collected by an image collection device of the vehicle, and a collection time corresponding to each road image and a positioning time corresponding to each pose information have a corresponding relationship; for each pose information, determining map semantic data corresponding to the pose information from the electronic navigation map; for each pose information, determining a quality detection result of the electronic navigation map based on map location information of the map semantic data corresponding to the pose information in the electronic navigation map, and for each pose information, determining a quality detection result of the electronic navigation map based on map location information of the map semantic data corresponding to the pose information in the electronic navigation map and observation location information of the perception data corresponding to the pose information in the corresponding road image; the perception data comprises perceived lane lines, and the map semantic data comprises map lane lines; the quality detection result comprises a detection result of a bending condition and / or a broken condition of the map lane lines; if the quality detection result comprises a detection result of a bending condition of the map lane lines, the step of determining, for each pose information, a quality detection result of the electronic navigation map based on map location information of the map semantic data corresponding to the pose information in the electronic navigation map comprises: for each pose information, determining mapping location information of the map lane line in a vehicle coordinate system corresponding to the pose information based on map location information of the map lane line in the map semantic data corresponding to the pose information in the electronic navigation map and the pose information, wherein each mapping location information comprises mapping location information of a plurality of discrete points corresponding to the corresponding map lane line; for each map lane line corresponding to each pose information, fitting a lane line fitting line corresponding to the map lane line based on mapping location information of a plurality of discrete points corresponding to the map lane line and a preset fitting algorithm; for each map lane line corresponding to each pose information, determining a distance variance corresponding to the map lane line based on the lane line fitting line corresponding to the map lane line and the mapping location information of the plurality of discrete points corresponding to the map lane line; if the distance variance exceeds a preset variance threshold, it is determined that the map lane line has a lane line bending condition; and / or, if the quality detection result comprises a detection result of a broken condition of the map lane lines, the step of determining, for each pose information, a quality detection result of the electronic navigation map based on map location information of the map semantic data corresponding to the pose information in the electronic navigation map and observation location information of the perception data corresponding to the pose information in the corresponding road image comprises: For each pose information, based on the map lane line corresponding to the pose information in the map semantic data corresponding to the pose information in the map position information of the electronic navigation map and the pose information, the conversion position information of the map lane line in the road image corresponding to the pose information is determined, wherein each conversion position information comprises conversion position information of a plurality of discrete points corresponding to the corresponding map lane line; For each map lane line corresponding to each pose information, based on the conversion position information of the plurality of discrete points corresponding to the map lane line, the distance between each adjacent two discrete points is calculated; If the distance between the adjacent two discrete points in the plurality of discrete points corresponding to the map lane line exceeds the preset difference value from the distance between the other adjacent two discrete points, it is determined that the map lane line has a lane line break.

2. The method of claim 1, wherein, The quality detection result includes the detection result of the position deviation of each map semantic data; The step of determining the quality detection result of the electronic navigation map based on the map position information of the electronic navigation map and the observation position information of the perception data corresponding to the pose information in the corresponding road image for each pose information, comprises: For each pose information, based on the map position information of the electronic navigation map and the pose information, the projection position information of the map semantic data in the road image corresponding to the pose information is determined. Based on the projection position information of the map semantic data in the road image corresponding to the pose information and the observation position information of the perception data corresponding to the map semantic data in the road image, the position deviation value between the map semantic data and the perception data corresponding thereto is determined. It is judged whether the position deviation value exceeds the preset distance threshold; If the judgment result is that the position deviation value exceeds the preset distance threshold, it is determined that the map semantic data in the electronic navigation map has a position deviation.

3. The method of claim 1, wherein, The perception data includes perception lane lines, and the map semantic data includes map lane lines; the quality detection result includes the detection result of the position deviation of the map lane line in the elevation direction; The step of determining the quality detection result of the electronic navigation map based on the map position information of the electronic navigation map and the observation position information of the perception data corresponding to the pose information in the corresponding road image for each pose information, comprises: For each pose information, based on the map lane line in the map semantic data corresponding to the pose information in the map position information of the electronic navigation map and the pose information, the mapping position information of the map lane line in the vehicle coordinate system corresponding to the pose information is determined. For each pose information, based on the observation position information of the perception lane line in the corresponding road image in the perception data corresponding to the pose information, the pose information and the projection matrix corresponding to the image acquisition device, the mapping position information of the perception lane line in the vehicle coordinate system corresponding to the pose information is determined. For each map lane corresponding to each pose information, based on the mapping position information of the map lane in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception lane corresponding to the map lane in the vehicle body coordinate system corresponding to the pose information, it is determined whether there is a position deviation of the map lane in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system. If it is determined that there is no position deviation of the map lane in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system, for each pose information, based on the map position information of the map lane in the electronic navigation map in the map semantic data corresponding to the pose information and the pose information, the projection position information of the map lane in the road image corresponding to the pose information is determined. For each map lane corresponding to each pose information, based on the projection position information of the map lane in the road image corresponding to the pose information and the observation position information of the perception lane in the perception data corresponding to the map lane in the road image, it is determined whether there is a position deviation of the map lane in the elevation direction.

4. The method of claim 3, wherein, The map lane includes a plurality of discrete points, and the mapping position information of the map lane in the vehicle body coordinate system corresponding to the pose information includes mapping position information of a plurality of discrete points corresponding to the map lane. For each map lane corresponding to each pose information, based on the mapping position information of the map lane in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception lane corresponding to the map lane in the vehicle body coordinate system corresponding to the pose information, it is determined whether there is a position deviation of the map lane in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system, comprising: For each discrete point included in each map lane corresponding to each pose information, based on the mapping position information of the discrete point and the mapping position information of the perception lane corresponding to the map lane in the vehicle body coordinate system corresponding to the pose information, the error sub-vector between the discrete point and the perception lane corresponding to the map lane is calculated. Based on the error sub-vector between each discrete point included in each map lane corresponding to each pose information and the perception lane corresponding to the map lane, the error vector between each map lane corresponding to each pose information and the perception lane corresponding to the map lane is determined as the error vector corresponding to each map lane. For each map lane corresponding to each pose information, based on the error vector corresponding to the map lane, the error vector component of the error vector corresponding to the map lane in the horizontal axis direction of the vehicle body coordinate system corresponding to the pose information and the error vector component in the vertical axis direction of the vehicle body coordinate system corresponding to the pose information are determined, and the modulus of the error vector component of the error vector corresponding to the map lane in the horizontal axis direction of the vehicle body coordinate system corresponding to the pose information and the modulus of the error vector component in the vertical axis direction of the vehicle body coordinate system corresponding to the pose information are calculated respectively. determining whether a module of an error vector component of the error vector corresponding to the map lane line in a lateral axis direction of the vehicle body coordinate system corresponding to the pose information exceeds a first value, if it is determined that the module of the error vector component of the error vector corresponding to the map lane line in the lateral axis direction of the vehicle body coordinate system corresponding to the pose information does not exceed the first value, it is determined that there is no position deviation of the map lane line in the lateral axis direction of the vehicle body coordinate system, if it is determined that the module of the error vector component of the error vector corresponding to the map lane line in the lateral axis direction of the vehicle body coordinate system corresponding to the pose information exceeds the first value, it is determined that there is a position deviation of the map lane line in the lateral axis direction of the vehicle body coordinate system; determining whether a module of an error vector component of the error vector corresponding to the map lane line in a longitudinal axis direction of the vehicle body coordinate system corresponding to the pose information exceeds a second value, if it is determined that the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle body coordinate system corresponding to the pose information does not exceed the second value, it is determined that there is no position deviation of the map lane line in the longitudinal axis direction of the vehicle body coordinate system, if it is determined that the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle body coordinate system corresponding to the pose information exceeds the first value, it is determined that there is a position deviation of the map lane line in the longitudinal axis direction of the vehicle body coordinate system.

5. The method according to any one of claims 1 to 4, characterized in that, The perception data includes a perception traffic sign, and the map semantic data includes a map traffic sign; and the quality detection result includes a detection result of a flip of the map traffic sign; The step of determining, for each pose information, a quality detection result of the electronic navigation map based on map position information of map semantic data corresponding to the pose information in the electronic navigation map and observation position information of perception data corresponding to the pose information in a corresponding road image, includes: For each pose information, determining, based on a map traffic sign in the map semantic data corresponding to the pose information and the pose information, mapping position information of the map traffic sign in a vehicle body coordinate system corresponding to the pose information; For each pose information, determining, based on observation position information of a perception traffic sign in a corresponding road image in the perception data corresponding to the pose information, the pose information, and a projection matrix corresponding to the image acquisition device, mapping position information of the perception traffic sign in a vehicle body coordinate system corresponding to the pose information; For each map traffic sign corresponding to a pose information, determining, based on mapping position information of the map traffic sign in a vehicle body coordinate system corresponding to the pose information and mapping position information of a perception traffic sign corresponding to the map traffic sign in a vehicle body coordinate system corresponding to the pose information, whether the map traffic sign is flipped.

6. The method of claim 5, wherein, The step of determining whether the map traffic sign is in a flipped state based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perceived traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information comprises: fitting a first fitting plane corresponding to the map traffic sign based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information and a preset plane fitting algorithm; fitting a second fitting plane corresponding to the perceived traffic sign corresponding to the map traffic sign based on the mapping position information of the perceived traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information and the preset plane fitting algorithm; calculating the included angle between the normal vector corresponding to the first fitting plane and the normal vector corresponding to the second fitting plane; if the included angle exceeds a preset angle, it is determined that the map traffic sign is in a flipped state.

7. The method of claim 6, wherein, After the steps of determining the quality detection result of the electronic navigation map based on the map position information of the map semantic data corresponding to each pose information in the electronic navigation map, and determining the quality detection result of the electronic navigation map based on the map position information of the map semantic data corresponding to each pose information in the electronic navigation map and the observation position information of the perceived data corresponding to the pose information in the corresponding road image, the method further comprises: if the quality detection result of the electronic navigation map indicates that the electronic navigation map has a quality problem, storing an abnormal statistical document corresponding to the electronic navigation map and abnormal image information, wherein the abnormal statistical document at least includes the identification of the map semantic data with a quality problem in the electronic navigation map and the problem quality type thereof, the map position information in the electronic navigation map, and the abnormal image information includes image information corresponding to the map semantic data with a quality problem.

8. An electronic navigation map quality detection method, characterized by, comprises: obtaining an electronic navigation map to be detected; obtaining trajectory information of a vehicle during driving in a scene corresponding to the electronic navigation map, wherein the trajectory information comprises a plurality of pose information and a positioning time corresponding to each pose information; obtaining perceived data determined by the vehicle during the driving, wherein each perceived data is data detected from a road image collected by an image collection device of the vehicle, and the collection time corresponding to each road image has a corresponding relationship with the positioning time corresponding to each pose information; for each pose information, determining map semantic data corresponding to the pose information from the electronic navigation map; for each pose information, determining a quality detection result of the electronic navigation map based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map and the observation position information of the perceived data corresponding to the pose information in the corresponding road image; The perception data comprises a perception traffic sign, and the map semantic data comprises a map traffic sign; the quality detection result comprises a detection result of a flip of the map traffic sign; The step of determining the quality detection result of the electronic navigation map based on the map location information of the map semantic data corresponding to each pose information in the electronic navigation map and the observation location information of the perception data corresponding to the pose information in the corresponding road image comprises: For each pose information, the mapping location information of the map traffic sign in the vehicle coordinate system corresponding to the pose information is determined based on the map location information of the map traffic sign in the map semantic data corresponding to the pose information in the electronic navigation map and the pose information; For each pose information, the mapping location information of the perception traffic sign in the vehicle coordinate system corresponding to the pose information is determined based on the observation location information of the perception traffic sign in the corresponding road image in the perception data corresponding to the pose information, the pose information and the projection matrix corresponding to the image acquisition device; For each map traffic sign corresponding to each pose information, whether the map traffic sign has a flip is determined based on the mapping location information of the map traffic sign in the vehicle coordinate system corresponding to the pose information and the mapping location information of the perception traffic sign corresponding to the map traffic sign in the vehicle coordinate system corresponding to the pose information.

9. The method of claim 8, wherein, The quality detection result comprises a detection result of a position deviation of each map semantic data; The step of determining the quality detection result of the electronic navigation map based on the map location information of the map semantic data corresponding to each pose information in the electronic navigation map and the observation location information of the perception data corresponding to the pose information in the corresponding road image comprises: For each pose information, the projection location information of the map semantic data in the road image corresponding to the pose information is determined based on the map location information of the map semantic data in the electronic navigation map and the pose information; A position deviation value between the map semantic data and the perception data corresponding to the map semantic data is determined based on the projection location information of the map semantic data in the road image corresponding to the pose information and the observation location information of the perception data corresponding to the map semantic data in the road image; It is judged whether the position deviation value exceeds a preset distance threshold; If the result of the judgment is that the position deviation value exceeds the preset distance threshold, it is determined that the map semantic data in the electronic navigation map has a position deviation.

10. The method of claim 8, wherein, The perception data comprises a perception lane line, and the map semantic data comprises a map lane line; the quality detection result comprises a detection result of a position deviation of the map lane line in the elevation direction; The step of determining the quality detection result of the electronic navigation map based on the map semantic data corresponding to each pose information in the map position information of the electronic navigation map and the observation position information of the perception data corresponding to the pose information comprises: For each pose information, the mapping position information of the map lane line in the vehicle body coordinate system corresponding to the pose information is determined based on the map lane line in the map semantic data corresponding to the pose information and the pose information. For each pose information, the mapping position information of the perception lane line in the vehicle body coordinate system corresponding to the pose information is determined based on the observation position information of the perception lane line in the corresponding road image in the perception data corresponding to the pose information, the pose information and the projection matrix corresponding to the image acquisition device. For each map lane line corresponding to each pose information, whether there is a position deviation in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system is determined based on the mapping position information of the map lane line in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information. If it is determined that there is no position deviation in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system, for each pose information, the projection position information of the map lane line in the road image corresponding to the pose information is determined based on the map lane line in the map semantic data corresponding to the pose information and the pose information in the map position information of the electronic navigation map. For each map lane line corresponding to each pose information, whether there is a position deviation in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system is determined based on the mapping position information of the map lane line in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information.

11. The method of claim 10, wherein, The map lane line comprises a plurality of discrete points, and the mapping position information of the map lane line in the vehicle body coordinate system corresponding to the pose information comprises mapping position information of a plurality of discrete points corresponding to the map lane line. For each map lane line corresponding to each pose information, whether there is a position deviation in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system is determined based on the mapping position information of the map lane line in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information. For each discrete point included in each map lane line corresponding to each pose information, an error sub-vector between the discrete point and the perception lane line corresponding to the map lane line is calculated based on the mapping position information of the discrete point and the mapping position information of the perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information. determine, for each map lane line corresponding to each pose information, an error vector between the map lane line and a perception lane line corresponding to the map lane line as an error vector corresponding to the map lane line based on error sub-vectors between each discrete point included in the map lane line and the perception lane line corresponding to the map lane line; for each map lane line corresponding to each pose information, determine, based on the error vector corresponding to the map lane line, an error vector component of the error vector corresponding to the map lane line in a lateral axis direction of a vehicle coordinate system corresponding to the pose information and an error vector component of the error vector corresponding to the map lane line in a longitudinal axis direction of the vehicle coordinate system corresponding to the pose information, and calculate a module of the error vector component of the error vector corresponding to the map lane line in the lateral axis direction of the vehicle coordinate system corresponding to the pose information and a module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle coordinate system corresponding to the pose information, respectively; determine whether the module of the error vector component of the error vector corresponding to the map lane line in the lateral axis direction of the vehicle coordinate system corresponding to the pose information exceeds a first value, and if the module of the error vector component of the error vector corresponding to the map lane line in the lateral axis direction of the vehicle coordinate system corresponding to the pose information does not exceed the first value, determine that there is no position deviation of the map lane line in the lateral axis direction of the vehicle coordinate system, and if the module of the error vector component of the error vector corresponding to the map lane line in the lateral axis direction of the vehicle coordinate system corresponding to the pose information exceeds the first value, determine that there is a position deviation of the map lane line in the lateral axis direction of the vehicle coordinate system; determine whether the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle coordinate system corresponding to the pose information exceeds a second value, and if the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle coordinate system corresponding to the pose information does not exceed the second value, determine that there is no position deviation of the map lane line in the longitudinal axis direction of the vehicle coordinate system, and if the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle coordinate system corresponding to the pose information exceeds the first value, determine that there is a position deviation of the map lane line in the longitudinal axis direction of the vehicle coordinate system.

12. The method of claim 8, wherein, The step of determining whether the map traffic sign is in a flipped state based on the mapping position information of the map traffic sign in the vehicle coordinate system corresponding to the pose information and the mapping position information of the perception traffic sign corresponding to the map traffic sign in the vehicle coordinate system corresponding to the pose information comprises: fitting a first fitting plane corresponding to the map traffic sign based on the mapping position information of the map traffic sign in the vehicle coordinate system corresponding to the pose information and a preset plane fitting algorithm; fitting a second fitting plane corresponding to the perception traffic sign corresponding to the map traffic sign based on the mapping position information of the perception traffic sign corresponding to the map traffic sign in the vehicle coordinate system corresponding to the pose information and the preset plane fitting algorithm; calculate an included angle between a normal vector corresponding to the first fitted plane and a normal vector corresponding to the second fitted plane; if the included angle exceeds a preset angle, determine that the traffic sign of the map exists a flipping situation.

13. The method according to any one of claims 8 to 12, wherein, After the step of determining the quality detection result of the electronic navigation map based on the map semantic data corresponding to the pose information in the map position information of the electronic navigation map and the observation position information of the perception data corresponding to the pose information in the corresponding road image for each pose information, the method further comprises: if the quality detection result of the electronic navigation map indicates that the electronic navigation map has a quality problem, storing an abnormal statistical document corresponding to the electronic navigation map and abnormal image information, wherein the abnormal statistical document at least includes an identifier of the map semantic data with a quality problem in the electronic navigation map and a problem quality type thereof, map position information in the electronic navigation map, and the abnormal image information includes image information corresponding to the map semantic data with a quality problem.

14. An electronic navigation map quality detection device, characterized by The device comprises: a first obtaining module configured to obtain an electronic navigation map to be detected; a second obtaining module configured to obtain trajectory information of a vehicle in a scene corresponding to the electronic navigation map during driving, wherein the trajectory information includes a plurality of pose information and a positioning time corresponding to each pose information; a third obtaining module configured to obtain perception data determined by the vehicle during the driving, wherein each perception data is data detected from a road image collected by an image collection device of the vehicle, and a collection time corresponding to each road image has a corresponding relationship with the positioning time corresponding to each pose information; a first determining module configured to determine, for each pose information, map semantic data corresponding to the pose information from the electronic navigation map; a second determining module configured to determine, for each pose information, a quality detection result of the electronic navigation map based on map position information of the map semantic data corresponding to the pose information in the electronic navigation map and / or observation position information of the perception data corresponding to the pose information in the corresponding road image; The perception data includes perceived lane lines, and the map semantic data includes map lane lines; the quality detection result includes detection results of map lane line bending conditions and / or breakage conditions; if the quality detection result includes a detection result of a map lane line bending condition, the second determining module is specifically configured to determine, for each pose information, mapping position information of the map lane line in a vehicle coordinate system corresponding to the pose information based on the map position information of the map lane line in the map semantic data corresponding to the pose information in the electronic navigation map and the pose information, wherein each mapping position information includes mapping position information of a plurality of discrete points corresponding to the corresponding map lane line; for each map lane line corresponding to each pose information, a lane line fitting line corresponding to the map lane line is fitted based on the mapping position information of the plurality of discrete points corresponding to the map lane line and a preset fitting algorithm. For each map lane line corresponding to each pose information, distance variance of the map lane line is determined based on a lane line fitting line corresponding to the map lane line and mapping position information of a plurality of discrete points corresponding to the map lane line; If the distance variance exceeds a preset variance threshold, it is determined that the map lane line has a lane line bending condition; If the quality detection result includes a detection result of a map lane line breakage condition, the second determination module is specifically configured to, for each pose information, determine, based on map lane line in map semantic data corresponding to the pose information and the pose information, conversion position information of the map lane line in a road image corresponding to the pose information, wherein each conversion position information includes conversion position information of a plurality of discrete points corresponding to the map lane line; For each map lane line corresponding to each pose information, distance between each two adjacent discrete points is calculated based on conversion position information of the plurality of discrete points corresponding to the map lane line; If, among the plurality of discrete points corresponding to the map lane line, there are two adjacent discrete points whose distance exceeds a preset difference value from distances between other adjacent discrete points, it is determined that the map lane line has a lane line breakage condition.

15. The apparatus of claim 14, wherein, The quality detection result includes a detection result of position deviation of each map semantic data; The second determination module is specifically configured to, for each pose information, determine, based on map semantic data corresponding to the pose information and the pose information, projection position information of the map semantic data in a road image corresponding to the pose information; Based on the projection position information of the map semantic data in the road image corresponding to the pose information and observation position information of the perception data corresponding to the map semantic data in the road image, a position deviation value between the map semantic data and the perception data corresponding thereto is determined; It is determined whether the position deviation value exceeds a preset distance threshold; If the determination result is that the position deviation value exceeds the preset distance threshold, it is determined that the map semantic data in the electronic navigation map has a position deviation condition.

16. The apparatus of claim 14, wherein, The perception data includes a perception lane line, and the map semantic data includes a map lane line; the quality detection result includes a detection result of position deviation of the map lane line in an elevation direction; The second determination module is specifically configured to, for each pose information, determine, based on map semantic data corresponding to the pose information and the pose information, mapping position information of a map lane line in the map semantic data in a vehicle coordinate system corresponding to the pose information; For each pose information, mapping position information of a perception lane line in a vehicle coordinate system corresponding to the pose information is determined based on observation position information of the perception lane line in a corresponding road image in perception data corresponding to the pose information, the pose information, and a projection matrix corresponding to the image acquisition device; For each map lane line corresponding to each pose information, based on mapping position information of the map lane line in a vehicle body coordinate system corresponding to the pose information and mapping position information of a perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information, it is determined whether there is a position deviation of the map lane line in a horizontal axis direction and a vertical axis direction of the vehicle body coordinate system; If it is determined that there is no position deviation of the map lane line in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system, for each pose information, based on map position information of the map lane line in the electronic navigation map in the map semantic data corresponding to the pose information and the pose information, projection position information of the map lane line in a road image corresponding to the pose information is determined; For each map lane line corresponding to each pose information, based on projection position information of the map lane line in a road image corresponding to the pose information and observation position information of a perception lane line in the road image in perception data corresponding to the map lane line, it is determined whether there is a position deviation of the map lane line in an elevation direction.

17. The apparatus of claim 16, wherein, The map lane line includes a plurality of discrete points, and the mapping position information of the map lane line in the vehicle body coordinate system corresponding to the pose information includes mapping position information of a plurality of discrete points corresponding to the map lane line; The second determination module is specifically configured to, for each discrete point included in each map lane line corresponding to each pose information, based on mapping position information of the discrete point and mapping position information of a perception lane line corresponding to the map lane line in a vehicle body coordinate system corresponding to the pose information, calculate an error sub-vector between the discrete point and the perception lane line corresponding to the map lane line; Based on the error sub-vector between each discrete point included in each map lane line corresponding to each pose information and a perception lane line corresponding to the map lane line, an error vector between each map lane line corresponding to each pose information and the perception lane line corresponding to the map lane line is determined as an error vector corresponding to each map lane line; For each map lane line corresponding to each pose information, based on the error vector corresponding to the map lane line, an error vector component of the error vector corresponding to the map lane line in a horizontal axis direction of a vehicle body coordinate system corresponding to the pose information and an error vector component of the error vector corresponding to the map lane line in a vertical axis direction of the vehicle body coordinate system corresponding to the pose information are determined, and a module of the error vector component of the error vector corresponding to the map lane line in the horizontal axis direction of the vehicle body coordinate system corresponding to the pose information and a module of the error vector component of the error vector corresponding to the map lane line in the vertical axis direction of the vehicle body coordinate system corresponding to the pose information are calculated. determining whether a module of an error vector component of the error vector corresponding to the map lane line in a lateral axis direction of the vehicle body coordinate system corresponding to the pose information exceeds a first value, if it is determined that the module of the error vector component of the error vector corresponding to the map lane line in the lateral axis direction of the vehicle body coordinate system corresponding to the pose information does not exceed the first data, it is determined that there is no position deviation of the map lane line in the lateral axis direction of the vehicle body coordinate system, if it is determined that the module of the error vector component of the error vector corresponding to the map lane line in the lateral axis direction of the vehicle body coordinate system corresponding to the pose information exceeds the first data, it is determined that there is a position deviation of the map lane line in the lateral axis direction of the vehicle body coordinate system; determining whether a module of an error vector component of the error vector corresponding to the map lane line in a longitudinal axis direction of the vehicle body coordinate system corresponding to the pose information exceeds a second value, if it is determined that the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle body coordinate system corresponding to the pose information does not exceed the second data, it is determined that there is no position deviation of the map lane line in the longitudinal axis direction of the vehicle body coordinate system, if it is determined that the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle body coordinate system corresponding to the pose information exceeds the first data, it is determined that there is a position deviation of the map lane line in the longitudinal axis direction of the vehicle body coordinate system.

18. The apparatus of any one of claims 14-17, wherein, The perception data includes a perceived traffic sign, and the map semantic data includes a map traffic sign; and the quality detection result includes a detection result of a flip of the map traffic sign. The second determining module is specifically configured to, for each piece of pose information, determine, based on map position information of the map traffic sign in the map semantic data corresponding to the pose information and the pose information, mapping position information of the map traffic sign in a vehicle body coordinate system corresponding to the pose information. For each piece of pose information, based on observation position information of a perceived traffic sign in corresponding road images in the perception data corresponding to the pose information, the pose information, and a projection matrix corresponding to the image acquisition device, mapping position information of the perceived traffic sign in a vehicle body coordinate system corresponding to the pose information is determined. For the map traffic sign corresponding to each piece of pose information, based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perceived traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information, it is determined whether the map traffic sign has a flip.

19. The apparatus of claim 18, wherein, The second determining module is specifically configured to, based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information and a preset plane fitting algorithm, fit a first fitting plane corresponding to the map traffic sign; based on the mapping position information of the perceived traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information and the preset plane fitting algorithm, a second fitting plane corresponding to the perceived traffic sign corresponding to the map traffic sign is fitted; and the second determining module is specifically configured to, for each piece of pose information, determine, based on mapping position information of a map traffic sign in the map semantic data corresponding to the pose information and the pose information, mapping position information of the map traffic sign in a vehicle body coordinate system corresponding to the pose information. calculate an included angle between a normal vector corresponding to the first fitted plane and a normal vector corresponding to the second fitted plane; if the included angle exceeds a preset angle, determine that the map traffic sign exists a flipping situation.

20. The apparatus of claim 19, wherein, The device further comprises: a storage module configured to, after the quality detection result of the electronic navigation map is determined based on the map semantic data corresponding to each piece of pose information in the electronic navigation map and / or the observation position information of the perception data corresponding to each piece of pose information in the corresponding road image, store an abnormal statistical document and abnormal image information corresponding to the electronic navigation map if the quality detection result of the electronic navigation map indicates that the electronic navigation map has a quality problem, wherein the abnormal statistical document at least includes an identifier of the map semantic data having a quality problem in the electronic navigation map and a problem quality type thereof, map position information in the electronic navigation map, and the abnormal image information includes image information corresponding to the map semantic data having a quality problem.

21. An electronic navigation map quality detection device, characterized by The device comprises: a first obtaining module configured to obtain an electronic navigation map to be detected; a second obtaining module configured to obtain trajectory information of a vehicle in a scene corresponding to the electronic navigation map during driving, wherein the trajectory information includes a plurality of pieces of pose information and a positioning time corresponding to each piece of pose information; a third obtaining module configured to obtain perception data determined by the vehicle during the driving, wherein each piece of perception data is data detected from a road image collected by an image collection device of the vehicle, and a collection time corresponding to each road image and a positioning time corresponding to each piece of pose information have a corresponding relationship; a first determining module configured to, for each piece of pose information, determine map semantic data corresponding to the pose information from the electronic navigation map; a second determining module configured to, for each piece of pose information, determine a quality detection result of the electronic navigation map based on map position information of the map semantic data corresponding to the pose information in the electronic navigation map and / or observation position information of the perception data corresponding to the pose information in the corresponding road image; The perception data includes a perceived traffic sign, and the map semantic data includes a map traffic sign; and the quality detection result includes a detection result of a flipping situation of the map traffic sign. The second determining module is specifically configured to, for each piece of pose information, determine mapping position information of the map traffic sign in a vehicle coordinate system corresponding to the pose information based on the map position information of the map traffic sign in the map semantic data corresponding to the pose information in the electronic navigation map and the pose information; For each piece of pose information, determine mapping position information of the perceived traffic sign in a vehicle coordinate system corresponding to the pose information based on the observation position information of the perceived traffic sign in the perception data corresponding to the pose information in the corresponding road image, the pose information, and a projection matrix corresponding to the image collection device. For each map traffic sign corresponding to the pose information, based on the mapping position information of the map traffic sign in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception traffic sign corresponding to the map traffic sign in the vehicle body coordinate system corresponding to the pose information, whether the map traffic sign exists in a flip situation is determined.

22. The apparatus of claim 21, wherein, The quality detection result includes a detection result of a position deviation situation of each map semantic data; The second determination module is specifically configured to, for each pose information, based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map and the pose information, determine projection position information of the map semantic data in a road image corresponding to the pose information; Based on the projection position information of the map semantic data in the road image corresponding to the pose information and the observation position information of the perception data corresponding to the map semantic data in the road image, a position deviation value between the map semantic data and the perception data corresponding thereto is determined. Whether the position deviation value exceeds a preset distance threshold is judged. If the judgment result is that the position deviation value exceeds the preset distance threshold, it is determined that the map semantic data in the electronic navigation map exists in a position deviation situation.

23. The apparatus of claim 21, wherein, The perception data includes a perception lane line, and the map semantic data includes a map lane line; the quality detection result includes a detection result of a position deviation of the map lane line in an elevation direction; The second determination module is specifically configured to, for each pose information, based on the map position information of the map lane line in the map semantic data corresponding to the pose information in the electronic navigation map and the pose information, determine mapping position information of the map lane line in a vehicle body coordinate system corresponding to the pose information; For each pose information, based on the observation position information of the perception lane line in the corresponding road image, the pose information, and a projection matrix corresponding to the image acquisition device, mapping position information of the perception lane line in a vehicle body coordinate system corresponding to the pose information is determined; For each map lane line corresponding to each pose information, based on the mapping position information of the map lane line in the vehicle body coordinate system corresponding to the pose information and the mapping position information of the perception lane line corresponding to the map lane line in the vehicle body coordinate system corresponding to the pose information, whether the map lane line exists in a position deviation situation in a horizontal axis direction and a vertical axis direction of the vehicle body coordinate system is determined; If it is determined that the map lane line does not exist in a position deviation situation in the horizontal axis direction and the vertical axis direction of the vehicle body coordinate system, for each pose information, based on the map position information of the map lane line in the map semantic data corresponding to the pose information in the electronic navigation map and the pose information, projection position information of the map lane line in a road image corresponding to the pose information is determined. For each map lane corresponding to each pose information, based on projection position information of the map lane in a road image corresponding to the pose information and observation position information of a perception lane corresponding to the map lane in the road image, it is determined whether the map lane has a position deviation in the elevation direction.

24. The apparatus of claim 23, wherein, The map lane includes a plurality of discrete points, and the mapping position information of the map lane in the vehicle coordinate system corresponding to the pose information includes mapping position information of the plurality of discrete points corresponding to the map lane. The second determination module is specifically configured to, for each discrete point included in each map lane corresponding to each pose information, calculate an error sub-vector between the discrete point and a perception lane corresponding to the map lane based on mapping position information of the discrete point and mapping position information of the perception lane in the vehicle coordinate system corresponding to the pose information. Based on the error sub-vector between each discrete point included in each map lane corresponding to each pose information and a perception lane corresponding to the map lane, an error vector between each map lane corresponding to each pose information and the perception lane corresponding to the map lane is determined as an error vector corresponding to each map lane. For each map lane corresponding to each pose information, based on the error vector corresponding to the map lane, an error vector component of the error vector corresponding to the map lane in a lateral axis direction of the vehicle coordinate system corresponding to the pose information and an error vector component of the error vector corresponding to the map lane in a longitudinal axis direction of the vehicle coordinate system corresponding to the pose information are determined, and a module of the error vector component of the error vector corresponding to the map lane in the lateral axis direction of the vehicle coordinate system corresponding to the pose information and a module of the error vector component of the error vector corresponding to the map lane in the longitudinal axis direction of the vehicle coordinate system corresponding to the pose information are calculated respectively. It is determined whether the module of the error vector component of the error vector corresponding to the map lane in the lateral axis direction of the vehicle coordinate system corresponding to the pose information exceeds a first value. If it is determined that the module of the error vector component of the error vector corresponding to the map lane in the lateral axis direction of the vehicle coordinate system corresponding to the pose information does not exceed the first value, it is determined that the map lane does not have a position deviation in the lateral axis direction of the vehicle coordinate system. If it is determined that the module of the error vector component of the error vector corresponding to the map lane in the lateral axis direction of the vehicle coordinate system corresponding to the pose information exceeds the first value, it is determined that the map lane has a position deviation in the lateral axis direction of the vehicle coordinate system. determining whether a module of an error vector component of the error vector corresponding to the map lane line in a longitudinal axis direction of the vehicle coordinate system corresponding to the pose information exceeds a second value, if it is determined that the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle coordinate system corresponding to the pose information does not exceed the second data, it is determined that the map lane line does not exist in the position deviation situation in the longitudinal axis direction of the vehicle coordinate system, if it is determined that the module of the error vector component of the error vector corresponding to the map lane line in the longitudinal axis direction of the vehicle coordinate system corresponding to the pose information exceeds the first data, it is determined that the map lane line exists in the position deviation situation in the longitudinal axis direction of the vehicle coordinate system.

25. The apparatus of claim 21, wherein, The second determining module is specifically configured to fit a first fitting plane corresponding to the map traffic sign based on the mapping position information of the map traffic sign in the vehicle coordinate system corresponding to the pose information and a preset plane fitting algorithm. fit a second fitting plane corresponding to the perceived traffic sign corresponding to the map traffic sign based on the mapping position information of the perceived traffic sign corresponding to the map traffic sign in the vehicle coordinate system corresponding to the pose information and the preset plane fitting algorithm; calculate an included angle between a normal vector corresponding to the first fitting plane and a normal vector corresponding to the second fitting plane; if the included angle exceeds a preset angle, it is determined that the map traffic sign exists in the turning situation.

26. The apparatus of any one of claims 21-25, wherein, The device further includes: The storage module is configured to, after determining the quality detection result of the electronic navigation map based on the map position information of the map semantic data corresponding to the pose information in the electronic navigation map and / or the observation position information of the perception data corresponding to the pose information in the corresponding road image for each pose information, if the quality detection result of the electronic navigation map represents that the electronic navigation map exists in a quality problem, store an abnormal statistical document and abnormal image information corresponding to the electronic navigation map, wherein the abnormal statistical document at least includes an identifier of the map semantic data existing in the quality problem in the electronic navigation map and a problem quality type thereof, map position information in the electronic navigation map, and the abnormal image information includes image information corresponding to the map semantic data existing in the quality problem.

Images