High-precision map data detection methods, equipment, roadside units, and edge computing platforms

Abstract

The present disclosure provides a map data detection method, which relates to the field of artificial intelligence technology, and in particular to technical fields such as autonomous driving, assisted driving, and high-precision maps. The specific implementation scheme is: determining a processing graph related to a target frame, wherein the processing graph includes a first node corresponding to the target frame and a second node corresponding to a preset point, the first node and the second node are connected via a first edge, and the two first nodes are connected via a second edge; determining at least one target processing subgraph from the processing graph, wherein the target processing subgraph includes at least one first edge; fusing the global pose information corresponding to the first edge in the target processing subgraph with the target relative pose information corresponding to the second edge to obtain target pose information; and determining a detection result based on the target pose information, wherein the detection result is used to indicate whether there is a deviation in the global pose information. The present disclosure also provides a map data detection device, an electronic device, and a storage medium.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, particularly to autonomous driving, assisted driving, high-precision mapping, and other technical fields. More specifically, the present disclosure provides a map data detection method, apparatus, electronic device, storage medium, computer program product, roadside unit, and edge computing platform. Background Art

[0002] High-precision maps, also known as high-accuracy maps, can be used by autonomous vehicles. These maps contain precise vehicle location information and rich road element data, helping vehicles anticipate complex road conditions, such as slope, curvature, and heading, to better mitigate potential risks. With the advancement of artificial intelligence and high-precision mapping technologies, the application scenarios for autonomous driving and assisted driving technologies are continuously expanding. In both autonomous and assisted driving modes, HD maps can be used to determine the vehicle's position and facilitate driving control. Summary of the Invention

[0003] The present disclosure provides a map data detection method, device, electronic device, storage medium, computer program product, roadside unit and edge computing platform.

[0004] According to one aspect of the present disclosure, a map data detection method is provided, the method comprising: determining a processing graph related to a target frame, wherein the processing graph comprises a first node corresponding to the target frame and a second node corresponding to a preset point, the first node and the second node are connected via a first edge, and the two first nodes are connected via a second edge, the first edge is used to represent global pose information of the first node, and the second edge is used to represent target relative pose information between the two first nodes, and the target frame is related to point cloud data of a target area; determining at least one target processing subgraph from the processing graph, wherein the target processing subgraph comprises at least one first edge; fusing the global pose information corresponding to the first edge in the target processing subgraph with the target relative pose information corresponding to the second edge to obtain target pose information; and determining a detection result based on the target pose information, wherein the detection result is used to indicate whether there is a deviation in the global pose information.

[0005] According to another aspect of the present disclosure, a map data detection device is provided, which includes: a first determination module for determining a processing graph related to a target frame, wherein the processing graph includes a first node corresponding to the target frame and a second node corresponding to a preset point, the first node and the second node are connected via a first edge, and the two first nodes are connected via a second edge, the first edge is used to represent the global pose information of the first node, and the second edge is used to represent the target relative pose information between the two first nodes, and the target frame is related to point cloud data of a target area; a second determination module for determining at least one target processing subgraph from the processing graph, wherein the target processing subgraph includes at least one first edge; a fusion module for fusing the global pose information corresponding to the first edge in the target processing subgraph with the target relative pose information corresponding to the second edge to obtain target pose information; and a third determination module for determining a detection result based on the target pose information, wherein the detection result is used to indicate whether there is a deviation in the global pose information.

[0006] According to another aspect of the present disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method provided according to the present disclosure.

[0007] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided. The computer instructions are used to cause a computer to execute the method provided according to the present disclosure.

[0008] According to another aspect of the present disclosure, a computer program product is provided, comprising a computer program, which implements the method provided according to the present disclosure when executed by a processor.

[0009] According to another aspect of the present disclosure, a roadside unit is provided, comprising the electronic device provided by the present disclosure.

[0010] According to another aspect of the present disclosure, an edge computing platform is provided, including multiple edge computing units, and the edge computing units include the electronic devices provided by the present disclosure.

[0011] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The accompanying drawings are provided to facilitate a better understanding of the present invention and do not constitute a limitation of the present disclosure.

[0013] Figure 1is a schematic diagram of an exemplary system architecture to which the map data detection method and apparatus can be applied according to an embodiment of the present disclosure;

[0014] Figure 2 is a flowchart of a map data detection method according to an embodiment of the present disclosure;

[0015] Figure 3A is an exemplary schematic diagram of a process diagram according to one embodiment of the present disclosure;

[0016] Figure 3B is an exemplary schematic diagram of a target processing subgraph according to an embodiment of the present disclosure;

[0017] Figure 3C is an exemplary schematic diagram of another target processing subgraph according to an embodiment of the present disclosure;

[0018] Figure 4 is a block diagram of a map data detection device according to an embodiment of the present disclosure; and

[0019] Figure 5 is a block diagram of an electronic device to which a map data detection method can be applied according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0020] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0021] To generate a high-precision map, a data acquisition device can be used to collect data from a target area multiple times, generating point cloud data related to the target area. For example, the data acquisition device can be deployed on an autonomous vehicle or a non-autonomous vehicle. The vehicle can then drive within the target area to collect point cloud data.

[0022] The data acquisition device may include multiple sensors. Multiple sensors include, for example, laser radar (LiDAR) and cameras. The data of the target area can be collected by the laser radar and the camera, and the data can be sent to the server after preprocessing. For example, the laser radar emits a laser scanning beam. When the laser scanning beam encounters an object and is reflected back, it is received by the laser radar, completing one emission and reception of the laser scanning beam. In this way, a large amount of point cloud data can be continuously collected. In practical applications, the data acquisition device can preprocess the large amount of point cloud data collected above, such as screening / filtering, to obtain preprocessed point cloud data.

[0023] The amount of point cloud data is large, and in the actual processing process, it is processed in frames. Point cloud data corresponding to the same area may exist between two or more frames of point cloud data. Point cloud data corresponding to the same area can be spliced ​​together through point cloud stitching. In addition, the laser radar pose corresponding to each target point cloud data can also be obtained through point cloud stitching. Among them, the laser radar pose corresponding to each frame is based on the center of the laser radar as the coordinate origin. As long as the pose of the laser radar is known, the coordinates of each laser point in the scanned point cloud data can be converted into coordinate values ​​in the global coordinate system.

[0024] The initial pose value of the lidar can be obtained through an inertial measurement unit (IMU) and a global navigation satellite system (GNSS) device. For example, the trajectory calculated by the IMU and GNSS is used as the initial pose value. For example, the pose can include position and attitude angles. The position corresponds to the three-dimensional coordinates (x, y, z), and the attitude angles include the rotation angles about the three coordinate axes: heading angle, pitch angle, and roll angle.

[0025] The displacement component of the global pose in this initial value can have deviations. For example, if the trajectory described above includes a portion with elevation drift, this will cause a deviation in the displacement component. This deviation can lead to map delamination or deviation, and even create safety risks.

[0026] However, this deviation is difficult to detect. When the trajectory is locally smooth and the relative pose is accurate, this deviation is difficult to detect. In this case, manual deviation detection is possible, but manual methods are costly.

[0027] Figure 1 is a schematic diagram of an exemplary system architecture to which the map data detection method and apparatus can be applied according to an embodiment of the present disclosure. It should be noted that: Figure 1 The examples shown are merely examples of system architectures to which the embodiments of the present disclosure may be applied, to help those skilled in the art understand the technical content of the present disclosure, but do not mean that the embodiments of the present disclosure may not be used in other devices, systems, environments or scenarios.

[0028] like Figure 1 As shown, the system architecture 100 according to this embodiment may include sensors 101, 102, 103, a network 120, a server 130, and a road side unit (RSU) 140. The network 120 is used as a medium for providing communication links between the sensors 101, 102, 103 and the server 130. The network 120 may include various connection types, such as wired and / or wireless communication links, etc.

[0029] The sensors 101 , 102 , 103 may interact with the server 130 via the network 120 to receive or send messages, etc.

[0030] Sensors 101, 102, and 103 may be functional components integrated into vehicle 110, such as infrared sensors, ultrasonic sensors, millimeter-wave radars, information collection devices, and the like. Sensors 101, 102, and 103 may be used to collect status data of perceived objects (e.g., pedestrians, vehicles, obstacles, etc.) around vehicle 110, as well as surrounding road data.

[0031] The vehicle 110 can communicate with the roadside unit 140, receive information from the roadside unit 140, or send information to the roadside unit.

[0032] The roadside unit 140 may be deployed on a traffic light, for example, to adjust the duration or frequency of the traffic light.

[0033] The server 130 may be located at a remote location capable of establishing communication with the vehicle-mounted terminal, and may be implemented as a distributed server cluster consisting of multiple servers, or as a single server.

[0034] Server 130 can be a server that provides various services. For example, mapping applications and data processing applications can be installed on server 130. For example, server 130 running a data processing application can receive obstacle status data and map data transmitted from sensors 101, 102, and 103 via network 120. One or more of the obstacle status data and map data can be used as data to be processed. The data to be processed is then processed to obtain target data.

[0035] It should be noted that the map data detection method provided in the embodiments of the present disclosure can generally be executed by server 130. Accordingly, the map data detection device provided in the embodiments of the present disclosure can also be located in server 130. However, this is not a limitation. The map data detection method provided in the embodiments of the present disclosure can also generally be executed by sensor 101, 102, or 103. Accordingly, the map data detection device provided in the embodiments of the present disclosure can also be located in sensor 101, 102, or 103.

[0036] I understand. Figure 1 The number of sensors, networks, and servers in the embodiment is only illustrative. Any number of sensors, networks, and servers may be used depending on the implementation requirements.

[0037] It should be noted that the sequence numbers of the operations in the following method are only used to indicate the operation for the purpose of description, and should not be regarded as indicating the order in which the operations should be performed. Unless explicitly stated, the method does not need to be performed in the order shown.

[0038] Figure 2 is a flowchart of a map data detection method according to an embodiment of the present disclosure.

[0039] like Figure 2 As shown, the method 200 may include operations S210 to S240.

[0040] In operation S210 , a processing graph associated with a target frame is determined.

[0041] In the disclosed embodiment, the target frame is associated with point cloud data of the target area.

[0042] For example, the data acquisition device described above can be used to acquire data from the target area to obtain multiple frames of point cloud data. According to preset filtering conditions, multiple target frames are filtered out from the multiple frames of point cloud data.

[0043] In an embodiment of the present disclosure, a processing graph includes a first node corresponding to a target frame and a second node corresponding to a preset point. The first node and the second node are connected via a first edge. Two first nodes are connected via a second edge. The first edge can represent global pose information of the first node. The second edge can represent relative pose information of the target between the two first nodes.

[0044] For example, the preset point may be the coordinate zero point of the global coordinate system.

[0045] For example, the processing graph can be a directed graph or an undirected graph.

[0046] In operation S220, at least one target processing subgraph is determined from the processing graph.

[0047] For example, the target processing subgraph includes at least one first edge.

[0048] For another example, the target processing subgraph may include the second edge and the first edge.

[0049] For example, the target processing subgraph may be a closed-loop graph formed by a first node, a second node, a first edge, and a second edge.

[0050] In operation S230 , the global pose information corresponding to the first edge in the target processing subgraph is fused with the target relative pose information corresponding to the second edge to obtain target pose information.

[0051] For example, the global pose information can correspond to a transformation matrix. The target relative pose information can also correspond to a transformation matrix. Various matrix operations are performed on the multiple transformation matrices to obtain operation results, thereby fusing the global pose information with the target relative pose information. For another example, the various matrix operations can include matrix multiplication operations.

[0052] In operation S240 , a detection result is determined based on the target pose information.

[0053] For example, the detection results are used to indicate whether there is a deviation in the global pose information.

[0054] For example, the target pose information may correspond to a transformation matrix, which may be the result of the operation described above.

[0055] For another example, the detection result may be determined based on a difference between the target pose information and the preset pose information. If the difference is greater than the preset difference, the detection result may be that there is a deviation in the global pose information.

[0056] Through the embodiments of the present disclosure, relative posture is used to detect whether there is a deviation in the global posture, so that when the true value is unknown, it can be accurately determined whether there is a deviation in the global posture, thereby reducing the cost required for deviation detection.

[0057] In some embodiments, the preset pose information may correspond to a unit matrix.

[0058] In some embodiments, the pose information corresponds to a transformation matrix.

[0059] For example, the relative pose information between two frames of point cloud data corresponds to a transformation matrix. Based on the transformation matrix, the relative rotation angle value and relative translation value between the two frames of point cloud data can be determined.

[0060] In some embodiments, the multiple target frames come from at least two point cloud data sets, and each point cloud data set in the at least two point cloud data sets is obtained by performing one data collection on the target area.

[0061] For example, the data acquisition device can collect data from the target area multiple times to obtain multiple point cloud data sets.

[0062] For another example, during each acquisition process, the data acquisition device may acquire a frame of data every 0.1 seconds, obtaining multiple initial frames. Key frames are then determined from the initial frames according to a first preset screening condition. The first preset screening condition may, for example, include: a relative rotation angle value greater than a preset relative rotation angle value, or a relative translation value greater than a first preset relative translation value. For example, the first preset relative translation value may be 10 meters, and the preset relative rotation angle value may be 10 degrees.

[0063] In one example, the first initial frame among the multiple initial frames can be determined as the first key frame. Based on the initial relative pose information between the first key frame and the other initial frames, multiple relative translation values ​​can be obtained, and the initial frame with a relative translation value of 11 meters relative to the first key frame is determined as the second key frame.

[0064] For another example, after multiple acquisitions, multiple key frames from multiple acquisition processes are obtained. Based on a second preset filtering condition, multiple target frames can be determined from the multiple key frames from different acquisition processes. The second preset filtering condition may, for example, include: a relative translation value being less than or equal to a second preset relative translation value. For example, the second preset relative translation value may be 30 meters.

[0065] In one example, taking 4 acquisitions as an example, after each acquisition, N key frames can be obtained, where N is an integer greater than or equal to 1. A key frame is randomly determined from the N key frames acquired for the first time as the target frame a. After determining the target frame a, multiple relative translation values ​​can be obtained based on the initial relative pose information between the target frame a and the N key frames acquired for the second time. Based on this, from the N key frames acquired for the second time, a key frame with a relative translation value of 30 meters with respect to the target frame a is determined as the target frame b. Next, the target frame c can be determined from the N key frames acquired for the third time, and the target frame d can be determined from the N key frames acquired for the fourth time.

[0066] In another example, taking 4 acquisitions as an example, N key frames can be obtained after each acquisition. A key frame is randomly determined from the N key frames acquired for the first time as the target frame a. After determining the target frame a, multiple relative translation values ​​can be obtained based on the initial relative pose information between the target frame a and the N key frames acquired for the second time. Based on this, from the N key frames acquired for the second time, a key frame with a relative translation value of 30 meters with respect to the target frame a is determined as the target frame b. The difference from the example described above is that after determining the target frame b, the target frame c can also be determined from the N key frames acquired for the first time.

[0067] Through the embodiments of the present disclosure, multiple target frames come from at least two point cloud data sets, which can accurately detect whether there is a deviation in the global posture.

[0068] In some embodiments, the key frame and the initial frame adjacent to the key frame are fused to obtain the target frame.

[0069] For example, taking four acquisitions as an example, each acquisition yields N keyframes. A keyframe a' is randomly selected from the N keyframes acquired in the first acquisition. Keyframe a' is fused with the three adjacent initial frames to yield the target frame a.

[0070] Through the embodiments of the present disclosure, the key frame is fused with the initial frame to add information to the sparse key frame, thereby enriching the relevant information in the target frame for more accurate detection.

[0071] In some embodiments, determining a processing graph includes: performing point cloud stitching on multiple target frames based on initial relative pose information of the multiple target frames to obtain target relative pose information between the multiple target frames; and determining a processing graph based on the target relative pose information between the multiple target frames.

[0072] For example, the initial relative pose information of the target frame may be determined during the process of acquiring the target frame.

[0073] For example, stitching can be performed based on the initial relative pose information, or the elevation-related components in the initial pose information can be set to zero before stitching.

[0074] It can be understood that the above determination of the processing Figure 1 Some embodiments have been described in detail, and the following will be combined with a detailed description of determining at least one target processing subgraph from a processing graph.

[0075] In some embodiments, determining at least one target processing subgraph from the processing graph includes: performing closed loop detection on the processing graph to obtain at least one initial processing subgraph; and determining the initial processing subgraph including the first edge as the target processing subgraph. Figures 3A to 3C Provide detailed explanation.

[0076] Figure 3A is an exemplary diagram of a process diagram according to one embodiment of the present disclosure.

[0077] like Figure 3AAs shown, processing graph 300 includes a first node 310, a first node 320, a first node 330, a first node 340, and a second node 350. First node 310 corresponds to target frame a, first node 320 corresponds to target frame b, first node 330 corresponds to target frame c, first node 340 corresponds to target frame d, and second node 350 corresponds to preset point G.

[0078] The first node 310 and the second node 350 are connected via a first edge 351. The first node 320 and the second node 350 are connected via a first edge 352. The first node 330 and the second node 350 are connected via a first edge 353. The first node 340 and the second node 350 are connected via a first edge 354.

[0079] The first node 310 and the first node 320 are connected via a second edge 312. The first node 310 and the first node 330 are connected via a second edge 313. The first node 310 and the first node 340 are connected via a second edge 314. The first node 320 and the first node 330 are connected via a second edge 323. The first node 320 and the first node 340 are connected via a second edge 324. The first node 330 and the first node 340 are connected via a second edge 334.

[0080] like Figure 3A As shown, the processing graph 300 is a directed graph. The data acquisition device collects point cloud data in a preset direction. Therefore, the direction of each edge in the processing graph 300 can be determined.

[0081] Figure 3B is an exemplary schematic diagram of a target processing subgraph according to an embodiment of the present disclosure. Figure 3C is an exemplary schematic diagram of another target processing subgraph according to an embodiment of the present disclosure.

[0082] In the embodiment of the present disclosure, closed-loop detection is performed on the processing graph 300 to obtain multiple initial processing subgraphs.

[0083] For example, an initial processing subgraph includes: first node 310, first node 320, second node 350, second edge 312, first edge 351, and first edge 352. The initial processing subgraph includes first edge 351 and first edge 352, and the initial processing subgraph can be used as a target processing subgraph 301.

[0084] For example, an initial processing subgraph includes: first node 310, first node 320, first node 340, and second node 350, second edge 312, second edge 324, first edge 351, and first edge 354. The initial processing subgraph includes first edge 351 and first edge 354, and the initial processing subgraph can be used as a target processing subgraph 302.

[0085] For another example, another processing subgraph includes the first node 310, the first node 320, the first node 330, the second edge 312, the second edge 313, and the second edge 323. This initial processing subgraph does not include the first edge. This initial processing subgraph may not be used as the target processing subgraph in this embodiment.

[0086] In some embodiments, the global pose information corresponding to the first edge in the target processing subgraph is fused with the target relative pose information corresponding to the second edge to obtain the target pose information, including: determining the target transformation matrix as the target pose information based on the transformation matrix corresponding to the target relative pose information and the transformation matrix corresponding to the global pose information.

[0087] In an embodiment of the present disclosure, determining the target transformation matrix includes: determining a preset direction of a directed edge in a target processing subgraph; determining a target directed edge opposite to the preset direction from a plurality of directed edges based on the directions of the plurality of directed edges and the preset directions of the directed edges; and determining the target transformation matrix based on the inverse matrix of the transformation matrix associated with the target directed edge and the transformation matrix associated with other directed edges except the target directed edge.

[0088] For example, a directed edge includes a first edge and a second edge.

[0089] For example, Figure 3B As shown, the target processing subgraph 301 may include: a first node 310, a first node 320, and a second node 350, a second edge 312, a first edge 351, and a first edge 352. The preset direction of the directed edges in the target processing subgraph may be, for example, a clockwise direction: from the first node 310 to the first node 320, from the first node 320 to the second node 350, and from the second node 350 to the first node 310.

[0090] like Figure 3B As shown, the direction of the first edge 352 is from the second node 350 to the first node 320. The direction of the first edge 352 is opposite to the preset direction. The first edge 352 can be used as a target directed edge.

[0091] like Figure 3BAs shown, the direction of first edge 351 is from second node 350 to first node 310. The direction of second edge 312 is from first node 310 to first node 320. The directions of first edge 351 and second edge 312 are the same as the preset directions. First edge 351 and second edge 312 can serve as other directed edges.

[0092] Second edge 312 may represent the relative pose information of the target between first node 310 and first node 320. For example, second edge 312 may represent the relative pose information of the target between target frame a and target frame b. This relative pose information may correspond to a transformation matrix M_ab. Transformation matrix M_ab may serve as a transformation matrix associated with second edge 312.

[0093] The first edge 351 may represent the global pose information of the first node 310 . The global pose information may correspond to a transformation matrix M_ga . The transformation matrix M_ga may serve as a transformation matrix associated with the first edge 351 .

[0094] The first edge 352 may represent the global pose information of the first node 320 . The global pose information may correspond to a transformation matrix M_gb . The transformation matrix M_gb may serve as a transformation matrix associated with the first edge 352 .

[0095] According to the transformation matrix M_ab, the inverse matrix of the transformation matrix M_gb (M_gb) -1 , transformation matrix M_ga, determine the target transformation matrix. For example, for the transformation matrix M_ab and the inverse matrix (M_gb) -1 Perform a matrix multiplication operation to obtain a first intermediate transformation matrix. Perform a matrix multiplication operation on the first intermediate transformation matrix and the transformation matrix M_ga to obtain a target transformation matrix M_abg.

[0096] For example, Figure 3C As shown, the target processing subgraph 302 may include: a first node 310, a first node 320, a first node 340, and a second node 350, a second edge 312, a second edge 324, a first edge 351, and a first edge 354. The preset direction of the directed edges in the target processing subgraph may be, for example, a clockwise direction: from the first node 310 to the first node 320, from the first node 320 to the first node 340, from the first node 340 to the second node 350, and from the second node 350 to the first node 310.

[0097] like Figure 3C As shown, the direction of the first edge 354 is from the second node 350 to the first node 340. The direction of the first edge 354 is opposite to the preset direction. The first edge 354 can be used as a target directed edge.

[0098] like Figure 3CAs shown, first edge 351 is oriented from second node 350 to first node 310. Second edge 312 is oriented from first node 310 to first node 320. Second edge 324 is oriented from first node 320 to first node 340. The directions of first edge 351, second edge 312, and second edge 324 are the same as the preset directions. First edge 351, second edge 312, and second edge 324 can serve as other directed edges.

[0099] Second edge 312 may represent the relative pose information of the target between first node 310 and first node 320. For example, second edge 312 may represent the relative pose information of the target between target frame a and target frame b. This relative pose information may correspond to a transformation matrix M_ab. Transformation matrix M_ab may serve as a transformation matrix associated with second edge 312.

[0100] Second edge 324 may represent the relative pose information of the target between first node 320 and first node 340. For example, second edge 324 may represent the relative pose information of the target between target frame b and target frame d. This relative pose information may correspond to a transformation matrix M_bd. Transformation matrix M_bd may serve as the transformation matrix associated with second edge 324.

[0101] The first edge 351 may represent the global pose information of the first node 310 . The global pose information may correspond to a transformation matrix M_ga . The transformation matrix M_ga may serve as a transformation matrix associated with the first edge 351 .

[0102] The first edge 354 may represent the global pose information of the first node 340 . The global pose information may correspond to a transformation matrix Mgd . The transformation matrix Mgd may serve as a transformation matrix associated with the first edge 354 .

[0103] According to the transformation matrix M_ab, transformation matrix M_bd, and the inverse matrix of transformation matrix M_gd (M_gd) -1 , transformation matrix M_ga, determine the target transformation matrix. For example, perform matrix multiplication on the transformation matrix M_ab and the transformation matrix M_bd to obtain the second intermediate transformation matrix. The second intermediate transformation matrix and the inverse matrix (M_gd) -1 Perform a matrix multiplication operation to obtain a third intermediate transformation matrix. Perform a matrix multiplication operation on the third intermediate transformation matrix and the transformation matrix M_ga to obtain a target transformation matrix M_abdg.

[0104] In an embodiment of the present disclosure, determining the detection result based on the target posture information includes: determining the translation value and the rotation angle value based on the target transformation matrix; determining the detection result of the first side based on the translation value, the rotation angle value, the target translation value and the target rotation angle value; and determining the detection result of the map data related to the target area based on the detection result of the first side.

[0105] For example, the translation value T_abg and the rotation angle value R_abg can be determined based on the target transformation matrix M_abg. For another example, the translation value T_abdg and the rotation angle value R_abdg can be determined based on the target transformation matrix M_abdg.

[0106] In an embodiment of the present disclosure, determining the detection result of the first edge based on the translation value, the rotation angle value, the target translation value and the target rotation angle value also includes: determining the edge quantity values ​​of the first edge and the second edge in the target processing subgraph; determining the target translation value and the target rotation angle value based on the edge quantity value, the preset translation value and the preset rotation angle value.

[0107] For example, Figure 3B As shown, target processing subgraph 301 may include second edge 312, first edge 351, and first edge 352. The number of edges in target processing subgraph 301 is 3. For example, if the preset translation value is 0.1 meter and the preset rotation angle value is 0.1 degree, for target processing subgraph 301, the target translation value may be 0.3 meter, and the target rotation angle value may be 0.3 degree.

[0108] For example, Figure 3C As shown, target processing subgraph 302 may include second edge 312, second edge 324, first edge 351, and first edge 354. The number of edges in target processing subgraph 302 is 4. For example, if the preset translation value is 0.1 meter and the preset rotation angle value is 0.1 degree, for target processing subgraph 302, the target translation value may be 0.4 meter, and the target rotation angle value may be 0.4 degree.

[0109] Next, in the embodiment of the present disclosure, it is possible to determine whether the translation value is less than or equal to the target translation value and whether the rotation angle value is less than or equal to the target rotation angle value to obtain a sub-detection result of the first edge. The sub-detection results include a first type of sub-detection result and a second type of sub-detection result.

[0110] For example, in response to determining that the translation value is less than or equal to the target translation value and the rotation angle value is less than or equal to the target rotation angle value, the sub-detection result of the first edge in the target processing subgraph is determined as a first-type sub-detection result. The first-type sub-detection result is used to indicate that there is no deviation in the global pose information corresponding to the first edge.

[0111] For another example, in response to determining that the translation value is greater than the target translation value or the rotation angle value is greater than the target rotation angle value, the sub-detection result for the first edge in the target processing subgraph is determined to be a second-type sub-detection result. The second-type sub-detection result is used to indicate that there is a deviation in the global pose information corresponding to the first edge.

[0112] In an example, if it is determined that the translation value T_abg is less than 0.3 meters and the rotation angle value R_abg is less than 0.3 degrees, a sub-detection result of the first edge 351 can be determined as a first-category sub-detection result, and a sub-detection result of the first edge 352 can also be determined as a first-category sub-detection result.

[0113] In an example, if it is determined that the translation value T_abdg is less than 0.4 meters and the rotation angle value R_abdg is less than 0.4 degrees, another sub-detection result of the first edge 351 can be determined as a first-category sub-detection result, and a sub-detection result of the first edge 354 can also be determined as a first-category sub-detection result.

[0114] In an embodiment of the present disclosure, determining the detection result of the first edge includes: determining at least one sub-detection result of the first edge based on at least one target processing subgraph related to the first edge; and determining the detection result of the first edge based on the at least one sub-detection result.

[0115] For example, Figure 3B and Figure 3C As shown, for first edge 351, the target processing subgraph associated with first edge 351 includes target processing subgraph 301, target processing subgraph 302, and so on. Among the multiple sub-detection results for first edge 351, if the number of sub-detection results of the first category is greater than the number of sub-detection results of the second category, the obtained detection result for the first edge can indicate that there is no deviation in the global pose information.

[0116] For another example, based on at least one of the detection results of the first edge 351, the detection result of the first edge 352, the detection result of the first edge 353, and the detection result of the first edge 354 indicating that there is a deviation in the global pose information, it can be determined that there is a global pose deviation in the point cloud data of the target area.

[0117] Through the embodiments of the present disclosure, when the processing graph is a directed graph, the detection efficiency can be improved, the amount of calculation can be reduced, and the cost of map data detection can be further reduced.

[0118] It can be understood that the processing graph 300 described above is a directed graph. In the embodiment of the present disclosure, the processing graph can also be an undirected graph, which will be described in detail below.

[0119] For example, when the processing graph is an undirected graph, at least one target processing subgraph may be determined from the processing graph. The target processing subgraph includes at least one first edge and at least one second edge.

[0120] For a target processing subgraph, a transformation matrix corresponding to at least one first edge and a transformation matrix corresponding to at least one second edge are obtained to obtain multiple transformation matrices. The inverse matrix of at least one of the multiple transformation matrices is obtained. Matrix multiplication is performed on the inverse matrix and the other transformation matrices to obtain a result. For multiple transformation matrices, multiple results can be obtained by performing operations using the inverse matrix of each transformation matrix. Among these results, the one with the smallest difference from the identity matrix is ​​used as the target pose information. Based on this target pose information, a detection result is determined.

[0121] Figure 4 is a block diagram of a map detection device according to an embodiment of the present disclosure.

[0122] like Figure 4 As shown, the apparatus 400 may include a first determination module 410 , a second determination module 420 , a fusion module 430 and a third determination module 440 .

[0123] The first determination module 410 is configured to determine a processing graph associated with the target frame. For example, the processing graph includes a first node corresponding to the target frame and a second node corresponding to a preset point. The first node and the second node are connected via a first edge, and two first nodes are connected via a second edge. The first edge is used to represent global pose information of the first node, and the second edge is used to represent relative pose information of the target between the two first nodes. The target frame is associated with point cloud data of the target area.

[0124] The second determining module 420 is configured to determine at least one target processing subgraph from the processing graph. For example, the target processing subgraph includes at least one first edge.

[0125] The fusion module 430 is used to fuse the global pose information corresponding to the first edge in the target processing subgraph with the target relative pose information corresponding to the second edge to obtain the target pose information.

[0126] The third determination module 440 is configured to determine a detection result based on the target posture information, wherein the detection result is used to indicate whether there is a deviation in the global posture information.

[0127] In some embodiments, the first determination module includes: a point cloud stitching module, which is used to perform point cloud stitching on multiple target frames based on the initial relative pose information of the multiple target frames to obtain target relative pose information between the multiple target frames; and a first determination submodule, which is used to determine a processing map based on the target relative pose information between the multiple target frames.

[0128] In some embodiments, the multiple target frames come from at least two point cloud data sets, and each point cloud data set in the at least two point cloud data sets is obtained by performing one data collection on the target area.

[0129] In some embodiments, the second determination module includes: a closed-loop detection submodule, configured to perform closed-loop detection on the processing graph to obtain at least one initial processing subgraph; and a second determination submodule, configured to determine the initial processing subgraph including the first edge as the target processing subgraph.

[0130] In some embodiments, the fusion module includes: a third determination submodule, used to determine the target transformation matrix as the target pose information based on the transformation matrix corresponding to the target relative pose information and the transformation matrix corresponding to the global pose information.

[0131] In some embodiments, the processing graph is a directed graph, and the third determination submodule includes: a first determination unit, used to determine the preset direction of the directed edges in the target processing subgraph, wherein the directed edges include a first edge and a second edge; a second determination unit, used to determine a target directed edge opposite to the preset direction from a plurality of directed edges based on the directions of the plurality of directed edges and the preset directions of the directed edges; and a third determination unit, used to determine the target transformation matrix based on the inverse matrix of the transformation matrix related to the target directed edge and the transformation matrix related to other directed edges except the target directed edge.

[0132] In some embodiments, the third determination module includes: a fourth determination submodule, used to determine the translation value and the rotation angle value based on the target transformation matrix; a fifth determination submodule, used to determine the detection result of the first edge based on the translation value, the rotation angle value, the target translation value and the target rotation angle value; and a sixth determination submodule, used to determine the detection result of the map data related to the target area based on the detection result of the first edge.

[0133] In some embodiments, the fifth determination submodule includes: a fourth determination unit, used to determine at least one sub-detection result of the first edge based on at least one target processing subgraph related to the first edge, wherein the sub-detection results include a first type of sub-detection result and a second type of sub-detection result, the first type of sub-detection result is used to indicate that there is no deviation in the global pose information corresponding to the first edge, and the second type of sub-detection result is used to indicate that there is a deviation in the global pose information corresponding to the first edge; and a fifth determination unit, used to determine the detection result of the first edge based on at least one sub-detection result.

[0134] In some embodiments, the fifth determination submodule includes: a sixth determination unit, which is used to determine the sub-detection result of the first edge in the target processing subgraph as a first-category sub-detection result in response to determining that the translation value is less than or equal to the target translation value and the rotation angle value is less than or equal to the target rotation angle value.

[0135] In some embodiments, the fifth determination submodule includes: a seventh determination unit, which is used to determine the sub-detection result of the first edge in the target processing subgraph as a second-category sub-detection result in response to determining that the translation value is greater than the target translation value or the rotation angle value is greater than the target rotation angle value.

[0136] In some embodiments, the fifth determination submodule also includes: an eighth determination unit, used to determine the edge quantity values ​​of the first edge and the second edge in the target processing subgraph; and a ninth determination unit, used to determine the target translation value and the target rotation angle value based on the edge quantity value, the preset translation value and the preset rotation angle value.

[0137] In the technical solutions disclosed herein, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0138] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, a computer program product, a roadside unit and an edge computing platform.

[0139] Figure 5 A schematic block diagram of an example electronic device 500 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are provided as examples only and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0140] like Figure 5 As shown, the device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. Various programs and data required for the operation of the device 500 can also be stored in the RAM 503. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0141] Various components in device 500 are connected to I / O interface 505, including: an input unit 506, such as a keyboard, mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a magnetic disk, optical disk, etc.; and a communication unit 509, such as a network card, modem, wireless communication transceiver, etc. The communication unit 509 allows device 500 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0142] The computing unit 501 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 501 performs the various methods and processes described above, such as the map data detection method. For example, in some embodiments, the map data detection method can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 508. In some embodiments, part or all of the computer program can be loaded and / or installed onto the device 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the map data detection method described above can be performed. Alternatively, in other embodiments, the computing unit 501 can be configured to perform the map data detection method by any other suitable means (e.g., via firmware).

[0143] Various embodiments of the systems and techniques described above can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0144] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0145] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0146] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0147] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0148] Computer systems may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The client and server relationship arises through computer programs running on the respective computers and having a client-server relationship to each other.

[0149] According to an embodiment of the present disclosure, the present disclosure further provides a roadside unit, which may include the electronic device provided by the present disclosure. For example, the roadside unit may include the electronic device 500 described above.

[0150] According to an embodiment of the present disclosure, the present disclosure further provides an edge computing platform, including multiple edge computing units, which may include the electronic device provided by the present disclosure. For example, the edge computing unit may include the electronic device 500 described above.

[0151] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not limited herein.

[0152] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A map data detection method, comprising: performing point cloud stitching on the plurality of target frames according to initial relative pose information of the plurality of target frames to obtain target relative pose information between the plurality of target frames; Determining a processing graph related to the target frame based on target relative pose information between the plurality of target frames, wherein the processing graph includes a first node corresponding to the target frame and a second node corresponding to a preset point, the first node and the second node are connected via a first edge, and two of the first nodes are connected via a second edge, the first edge is used to represent global pose information of the first node, and the second edge is used to represent target relative pose information between two of the first nodes, and the target frame is related to point cloud data of a target area; Performing closed-loop detection on the processing graph to obtain at least one initial processing subgraph; determining at least one of the initial processing subgraphs including the first edge as at least one target processing subgraph, wherein the target processing subgraph includes at least one of the first edges; In the target processing subgraph, determining a target transformation matrix as target pose information according to a transformation matrix corresponding to the target relative pose information and a transformation matrix corresponding to the global pose information; and A detection result is determined based on the target posture information, wherein the detection result is used to indicate whether there is a deviation in the global posture information.

2. The method according to claim 1, wherein The plurality of target frames come from at least two point cloud data sets, and each of the at least two point cloud data sets is obtained by performing one data collection on the target area.

3. The method according to claim 1, wherein The processing graph is a directed graph, Determining the target transformation matrix includes: Determining a preset direction of a directed edge in the target processing subgraph, wherein the directed edge includes the first edge and the second edge; According to the directions of the plurality of directed edges and the preset directions of the directed edges, determining a target directed edge having a direction opposite to the preset direction from the plurality of directed edges; and The target transformation matrix is ​​determined according to an inverse matrix of the transformation matrix associated with the target directed edge and transformation matrices associated with other directed edges except the target directed edge.

4. The method according to claim 1, wherein Determining the detection result according to the target posture information includes: Determine the translation value and the rotation angle value according to the target transformation matrix; determining a detection result of the first edge according to the translation value, the rotation angle value, the target translation value, and the target rotation angle value; and A detection result of map data related to the target area is determined according to the detection result of the first side.

5. The method according to claim 4, wherein Determining the detection result of the first edge includes: Determining, based on at least one target processing subgraph associated with the first edge, at least one sub-detection result of the first edge, wherein the sub-detection results include a first type of sub-detection result and a second type of sub-detection result, the first type of sub-detection result being used to indicate that global pose information corresponding to the first edge does not have a deviation, and the second type of sub-detection result being used to indicate that global pose information corresponding to the first edge has a deviation; and A detection result of the first edge is determined according to at least one of the sub-detection results.

6. The method according to claim 5, wherein: Determining the detection result of the first edge according to the translation value, the rotation angle value, the target translation value, and the target rotation angle value includes: In response to determining that the translation value is less than or equal to the target translation value and the rotation angle value is less than or equal to the target rotation angle value, the sub-detection result of the first edge in the target processing subgraph is determined as the first type of sub-detection result.

7. The method according to claim 5, wherein: Determining the detection result of the first edge according to the translation value, the rotation angle value, the target translation value, and the target rotation angle value includes: In response to determining that the translation value is greater than the target translation value or the rotation angle value is greater than the target rotation angle value, the sub-detection result of the first edge in the target processing subgraph is determined as the second-type sub-detection result.

8. The method according to claim 4, wherein Determining the detection result of the first edge according to the translation value, the rotation angle value, the target translation value, and the target rotation angle value further includes: Determine the edge quantity values ​​of the first edge and the second edge in the target processing subgraph; and The target translation value and the target rotation angle value are determined according to the edge quantity value, the preset translation value and the preset rotation angle value.

9. A map data detection device comprising: A point cloud stitching submodule is used to stitch point clouds of a plurality of target frames according to their initial relative pose information, so as to obtain target relative pose information between the plurality of target frames; a first determining submodule, configured to determine a processing graph associated with the target frame based on target relative pose information between the plurality of target frames, wherein the processing graph includes a first node corresponding to the target frame and a second node corresponding to a preset point, the first node and the second node are connected via a first edge, and two first nodes are connected via a second edge, the first edge is used to represent global pose information of the first node, the second edge is used to represent target relative pose information between two first nodes, and the target frame is associated with point cloud data of a target area; a closed-loop detection submodule, configured to perform closed-loop detection on the processing graph to obtain at least one initial processing subgraph; a second determining submodule, configured to determine at least one of the initial processing subgraphs including the first edge as at least one target processing subgraph, wherein the target processing subgraph includes at least one of the first edges; A third determining submodule is configured to determine, in the target processing subgraph, a target transformation matrix as the target pose information according to the transformation matrix corresponding to the target relative pose information and the transformation matrix corresponding to the global pose information; and The third determination module is used to determine a detection result based on the target posture information, wherein the detection result is used to indicate whether there is a deviation in the global posture information.

10. The device according to claim 9, wherein The plurality of target frames come from at least two point cloud data sets, and each of the at least two point cloud data sets is obtained by performing one data collection on the target area.

11. The device according to claim 9, wherein The processing graph is a directed graph, The third determining submodule includes: a first determining unit, configured to determine a preset direction of a directed edge in the target processing subgraph, wherein the directed edge includes the first edge and the second edge; a second determining unit, configured to determine, based on the directions of the plurality of directed edges and the preset directions of the directed edges, a target directed edge having a direction opposite to the preset direction from the plurality of directed edges; and The third determining unit is configured to determine the target transformation matrix according to an inverse matrix of the transformation matrix associated with the target directed edge and transformation matrices associated with other directed edges except the target directed edge.

12. The device according to claim 9, wherein The third determining module includes: A fourth determination submodule, configured to determine a translation value and a rotation angle value according to the target transformation matrix; a fifth determining submodule, configured to determine a detection result of the first edge according to the translation value, the rotation angle value, the target translation value, and the target rotation angle value; and The sixth determining submodule is configured to determine a detection result of the map data related to the target area according to the detection result of the first edge.

13. The device according to claim 12, wherein The fifth determining submodule includes: a fourth determining unit, configured to determine, based on at least one target processing subgraph associated with the first edge, at least one sub-detection result of the first edge, wherein the sub-detection results include a first type of sub-detection result and a second type of sub-detection result, the first type of sub-detection result being used to indicate that there is no deviation in the global pose information corresponding to the first edge, and the second type of sub-detection result being used to indicate that there is a deviation in the global pose information corresponding to the first edge; and A fifth determining unit is configured to determine a detection result of the first edge according to at least one of the sub-detection results.

14. The device according to claim 13, wherein The fifth determining submodule includes: A sixth determining unit is configured to determine the sub-detection result of the first edge in the target processing subgraph as the first type of sub-detection result in response to determining that the translation value is less than or equal to the target translation value and the rotation angle value is less than or equal to the target rotation angle value.

15. The device according to claim 13, wherein The fifth determining submodule includes: A seventh determining unit is configured to determine, in response to determining that the translation value is greater than the target translation value or the rotation angle value is greater than the target rotation angle value, the sub-detection result of the first edge in the target processing subgraph as the second-category sub-detection result.

16. The device according to claim 12, wherein The fifth determining submodule further includes: An eighth determining unit, configured to determine the edge quantity values ​​of the first edge and the second edge in the target processing subgraph; and A ninth determining unit is configured to determine the target translation value and the target rotation angle value according to the edge quantity value, the preset translation value, and the preset rotation angle value.

17. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 8.

18. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause a computer to execute the method according to any one of claims 1 to 8.

19. A computer program product comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 8.

20. A roadside unit comprising the electronic device according to claim 17.

21. An edge computing platform comprising a plurality of edge computing units, wherein the edge computing units include the electronic device according to claim 17.

Images