Vector map construction method and device, electronic equipment and medium

Abstract

The invention relates to a vector map construction method and device, electronic equipment and a medium, and particularly relates to the technical field of intelligent driving. Comprising the steps that a to-be-detected lane line and an original map are acquired, the to-be-detected lane line is obtained based on aerial view recognition of a current frame, and a target part of the lane line is marked in the original map; segmenting the target part of the lane line to be measured according to the original map to obtain a segmented curve of the current frame; calculating a correlation parameter between the segmented curve and the historical lane line, wherein the correlation parameter is used for representing the correlation possibility of the segmented curve and the historical lane line; and when the association parameter is greater than or equal to a preset association threshold value, associating the segmented curve with a historical lane line to obtain a vector map. The embodiment of the invention is used for improving the consistency and stability of the lane lines in the vector map.

Description

Technical Field

[0001] The present disclosure relates to the field of intelligent driving technology, and in particular to a method, device, electronic device and medium for constructing a vector map. Background Art

[0002] Vector maps are widely used in systems such as Geographic Information System (GIS), positioning systems, navigation systems, and autonomous driving systems. In related technologies, the absolute coordinates of point clouds in the world coordinate system are mainly obtained through sensors (such as lidar) and high-precision combined navigation equipment, and then objects of interest (such as lane lines, fences, traffic lights, signs, etc.) are identified based on the bird's-eye view, and vectorized calculations are performed one by one, and finally converted into a standard map format to generate a map containing real-time vector information. However, there are relatively complex lane lines in the actual road environment, such as ramps, large curvature curves, etc., and the identification of lane lines based on bird's-eye views is also easily interfered by factors such as environmental occlusion and lighting changes. These situations lead to the problem of sudden changes in the position, length, line type, and other elements of the same lane line in adjacent frames, which in turn causes the constructed vector map to have lane line jumps, poor continuity, and other unstable lane line problems; therefore, how to reduce the problem of unstable lane line generation in vector maps has become an urgent problem to be solved. Summary of the invention

[0003] In order to solve the above technical problems or at least partially solve the above technical problems, the present disclosure provides a method, device, electronic device and medium for constructing a vector map, which can improve the consistency and stability of the vector map between adjacent frames.

[0004] In order to achieve the above objectives, the technical solutions provided by the embodiments of the present disclosure are as follows:

[0005] In a first aspect, the present disclosure provides a method for constructing a vector map, comprising:

[0006] Obtaining a lane line to be tested and an original map, wherein the lane line to be tested is identified based on a bird's-eye view of the current frame, and a target portion of the lane line is marked in the original map;

[0007] Segmenting the target portion of the lane line to be tested according to the original map to obtain a segmented curve of the current frame;

[0008] Calculating an association parameter between the segmented curve and the historical lane line, wherein the association parameter is used to indicate a possibility of association between the segmented curve and the historical lane line;

[0009] When the association parameter is greater than or equal to a preset association threshold, the segmented curve is associated with the historical lane line to obtain a vector map.

[0010] As an optional implementation of the embodiment of the present disclosure, segmenting the target portion of the lane line to be tested according to the original map to obtain the segmented curve of the current frame includes:

[0011] Obtaining a curve equation corresponding to the lane line to be tested, and determining the curvatures of a plurality of target curves; wherein each of the target curves is a curve formed by a first preset number of adjacent points on the lane line to be tested;

[0012] Calculate the curvature difference between adjacent target curves;

[0013] When the curvature difference is greater than a preset curvature difference and the multiple target curves correspond to the target portion of the lane line to be measured, the target portion of the lane line to be measured is segmented according to the multiple target curves to obtain the segmented curves of the current frame.

[0014] As an optional implementation of the embodiment of the present disclosure, after segmenting the target portion of the lane line to be tested according to the original map to obtain the segmented curve of the current frame, and before calculating the association parameter between the segmented curve and the historical lane line, the method further includes:

[0015] Acquire historical lane lines, and project the historical lane lines into the vehicle coordinate system of the current frame;

[0016] The calculating the association parameter between the segmented curve and the historical lane line includes:

[0017] In the vehicle coordinate system of the current frame, an association parameter between the segmented curve and the historical lane line is calculated.

[0018] As an optional implementation of the embodiment of the present disclosure, obtaining historical lane lines and projecting the historical lane lines to the vehicle coordinate system of the current frame includes:

[0019] Obtain odometer information, and determine the vehicle's posture change information based on the odometer information;

[0020] According to the posture change information, the vehicle coordinate system of the historical frame corresponding to the historical lane line is converted into the vehicle coordinate system of the current frame, so as to project the historical lane line into the vehicle coordinate system of the current frame.

[0021] As an optional implementation of the embodiment of the present disclosure, calculating the association parameter between the segmented curve and the historical lane line includes:

[0022] Extracting a second preset number of visual points from the segmented curve;

[0023] Calculating a residual value between each of the visual points and the historical lane line;

[0024] Calculating a residual mean value according to the residual value of each visual point and the second preset number;

[0025] The reciprocal of the residual mean is used as the correlation parameter.

[0026] As an optional implementation of the embodiment of the present disclosure, when the association parameter is greater than or equal to a preset association threshold, the segmented curve is associated with the historical lane line to obtain a vector map, including:

[0027] When the association parameter is greater than or equal to a preset association threshold, determining a horizontal coordinate value of the visual point on the segmented curve, wherein the horizontal coordinate value is used to represent the distance between the visual point and the vehicle;

[0028] Setting a weight coefficient of the visual point according to the horizontal coordinate value, wherein the weight coefficient is in inverse proportion to the horizontal coordinate value;

[0029] Fitting is performed according to the historical lane lines, the segmented curves and the weight coefficients to obtain the vector map.

[0030] As an optional implementation of the embodiment of the present disclosure, fitting is performed according to the historical lane line, the segmented curve and the weight coefficient to obtain the vector map, including:

[0031] Performing fitting according to the historical lane line, the segmented curve and the weight coefficient;

[0032] The fitting result is filtered to remove noise in the fitting result to obtain the vector map.

[0033] In a second aspect, the present disclosure provides a device for constructing a vector map, the device comprising:

[0034] An acquisition module, used to acquire a lane line to be tested and an original map, wherein the lane line to be tested is identified based on a bird's-eye view of the current frame, and a target portion of the lane line is marked in the original map;

[0035] A segmentation module, used for segmenting the target part of the lane line to be tested according to the original map to obtain the segmentation curve of the current frame;

[0036] A calculation module, used for calculating an association parameter between the segmented curve and the historical lane line, wherein the association parameter is used for indicating a possibility of association between the segmented curve and the historical lane line;

[0037] The association module is used to associate the segmented curve with the historical lane line to obtain a vector map when the association parameter is greater than or equal to a preset association threshold.

[0038] As an optional implementation of the embodiment of the present disclosure, the segmentation module is specifically used to: obtain a curve equation corresponding to the lane line to be tested, and determine the curvatures of multiple target curves; wherein each of the target curves is a curve composed of a first preset number of adjacent points on the lane line to be tested;

[0039] Calculate the curvature difference between adjacent target curves;

[0040] When the curvature difference is greater than a preset curvature difference and the multiple target curves correspond to the target portion of the lane line to be measured, the target portion of the lane line to be measured is segmented according to the multiple target curves to obtain the segmented curves of the current frame.

[0041] As an optional implementation of the embodiment of the present disclosure, the calculation module is further used to: obtain historical lane lines, and project the historical lane lines to the vehicle coordinate system of the current frame;

[0042] The calculation module is specifically used to calculate the association parameters between the segmented curve and the historical lane line in the vehicle coordinate system of the current frame.

[0043] As an optional implementation of the embodiment of the present disclosure, the calculation module is specifically used to: obtain odometer information, and determine the posture change information of the vehicle according to the odometer information;

[0044] According to the posture change information, the vehicle coordinate system of the historical frame corresponding to the historical lane line is converted into the vehicle coordinate system of the current frame, so as to project the historical lane line into the vehicle coordinate system of the current frame.

[0045] As an optional implementation of the embodiment of the present disclosure, the calculation module is specifically used to: extract a second preset number of visual points from the piecewise curve;

[0046] Calculating a residual value between each of the visual points and the historical lane line;

[0047] Calculating a residual mean value according to the residual value of each visual point and the second preset number;

[0048] The reciprocal of the residual mean is used as the correlation parameter.

[0049] As an optional implementation of the embodiment of the present disclosure, the association module is specifically used to: determine the abscissa value of the visual point on the piecewise curve when the association parameter is greater than or equal to a preset association threshold, and the abscissa value is used to represent the distance between the visual point and the vehicle;

[0050] Setting a weight coefficient of the visual point according to the horizontal coordinate value, wherein the weight coefficient is in inverse proportion to the horizontal coordinate value;

[0051] Fitting is performed according to the historical lane lines, the segmented curves and the weight coefficients to obtain the vector map.

[0052] As an optional implementation of the embodiment of the present disclosure, the association module is specifically used to: perform fitting according to the historical lane line, the segmented curve and the weight coefficient;

[0053] The fitting result is filtered to remove noise in the fitting result to obtain the vector map.

[0054] In a third aspect, the present disclosure provides an electronic device comprising: a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the method for constructing a vector map as described in the first aspect or any one of its optional embodiments.

[0055] In a fourth aspect, the present disclosure provides a computer-readable storage medium, comprising: a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the method for constructing a vector map as described in the first aspect or any optional embodiment thereof is implemented.

[0056] In a fifth aspect, the present disclosure provides a computer program product, including: the computer program product includes a computer program, and when the computer program runs on a computer, the computer implements the vector map construction method as described in the first aspect or any optional embodiment thereof.

[0057] Compared with the prior art, the technical solution provided by the embodiments of the present disclosure has the following advantages:

[0058] The present disclosure provides a method, device, electronic device and medium for constructing a vector map, wherein the method first obtains the lane line to be tested identified based on the bird's-eye view of the current frame, and the original map with the target part of the lane line marked, and then segments the target part of the lane line to be tested according to the original map to obtain the segmented curve of the current frame, and further calculates the association parameter between the segmented curve and the historical lane line, the association parameter is used to indicate the possibility of the segmented curve being associated with the historical lane line; when the association parameter is greater than or equal to a preset association threshold, the segmented curve is associated with the historical lane line, thereby obtaining a vector map. The present disclosure segments the target part of the lane line to be tested corresponding to the current frame, and respectively determines the possibility of each segmented curve being associated with the historical lane line, thereby associating the segmented curve with a high possibility with the historical lane line, accurately identifying the same lane line in the current frame and the historical frame, and improving the consistency and stability of the lane lines in the vector map. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0060] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0061] Figure 1 A schematic diagram of a process of constructing a vector map according to an embodiment of the present disclosure;

[0062] Figure 2 A schematic diagram of the structure of a vector map construction device provided in an embodiment of the present disclosure;

[0063] Figure 3 A schematic structural diagram of an electronic device is provided in accordance with an embodiment of the present disclosure. DETAILED DESCRIPTION

[0064] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0065] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0066] A method for constructing a vector map provided in an embodiment of the present disclosure can be implemented by a computer device, including but not limited to vehicle-mounted devices, servers, personal computers, laptops, tablet computers, smart phones, etc. Computer devices include user devices and network devices. Among them, user devices include but are not limited to vehicle terminals, computers, smart phones, tablet computers, etc.; network devices include but are not limited to a single network server, a server group composed of multiple network servers, or a cloud composed of a large number of computers or network servers in cloud computing, wherein cloud computing is a type of distributed computing, a super virtual computer composed of a group of loosely coupled computer sets. Among them, the computer device can be operated alone to implement the present disclosure, and can also be connected to the network and implement the present disclosure through interactive operations with other computer devices in the network. Among them, the network where the computer device is located includes but is not limited to the Internet, wide area network, metropolitan area network, local area network, virtual private network (VPN) network, etc.

[0067] It should be noted that the protection scope of the vector map-based construction method described in the embodiment of the present disclosure is not limited to the execution order of the steps listed in this embodiment. All solutions implemented by adding, reducing or replacing steps in the prior art based on the principles of the present disclosure are included in the protection scope of the present disclosure.

[0068] like Figure 1 As shown, Figure 1 FIG. 1 is a flow chart of a method for constructing a vector map according to an embodiment of the present disclosure. The method may be executed by a device for constructing a vector map, wherein the device may be implemented by software and / or hardware and may generally be integrated in an electronic device. Figure 1 As shown, the method mainly includes the following steps S101 to S104:

[0069] S101. Obtain the lane line to be tested and the original map.

[0070] In the present disclosure, lane lines are described using cubic curves and start and end point positions in the vehicle coordinate system. For example, the feature vector of the lane line to be measured can be [c3, c2, c1, c0, start, end], where c3, c2, c1, c0 are cubic curves y = c3*x 3 +c2*x 2 +c1*x+c0 is the coefficient of the independent variable, start is the starting point of the lane line to be tested, and end is the end point of the lane line to be tested.

[0071] In some embodiments, the process of obtaining the lane line to be measured includes: first obtaining an image collected by the sensor at the current moment, for example, by using multiple fisheye cameras installed on the vehicle to capture the surrounding environment of the vehicle, and then fusing and splicing the captured images to obtain a bird's-eye view (Bird Eye View, BEV); then identifying and extracting the lane line to be measured in the bird's-eye view of the current frame from the bird's-eye view through a neural learning network. For specific identification and perception of lane lines from the bird's-eye view, reference may be made to the prior art, and this disclosure will not be elaborated on here.

[0072] The above embodiment identifies the bird's-eye view of the current frame at the current moment to obtain the lane line to be tested that needs to be associated. In this case, the lane line to be tested is relatively independent and has not been assigned to any historical lane line previously identified. Subsequent steps need to be performed to determine which historical lane line the lane line to be tested is the same lane line.

[0073] In some embodiments, the original map may be a Standard Definition (SD) map or a High Definition (HD) map. The target portion of the lane line is annotated in advance in the original map, and the target portion is the portion of the lane line that has a non-smooth transition. For example, the target portion of the lane line is the portion other than the straight lane, including but not limited to: the lane line in intersections, ramps, and large curvature curves. The target portion of the lane line may be manually annotated in advance in the original map, or may be identified and annotated in the original map by an artificial intelligence algorithm such as a neural network model, and the present disclosure does not specifically limit this.

[0074] S102: Segment the target part of the lane line to be measured according to the original map to obtain a segmented curve of the current frame.

[0075] In some embodiments, the target portion of the lane line to be tested is determined based on the prior annotations of the lane lines in the original map. The original map includes lane lines, and the target portions of all lane lines such as intersections, ramps, and large curvature curves have been marked. After step S101, the cubic curve of the lane line to be tested is obtained, and the cubic curve is divided into a cubic curve corresponding to the target portion and a cubic curve corresponding to other portions based on the prior annotations of the original map. For example, the ramp position of the lane line to be tested has been marked in the original map, and the lane line to be tested is divided into a ramp curve and other curves according to the ramp position of the lane line to be tested, and the ramp curve is used as the segmented curve of the current frame. In this way, the lane line to be tested is segmented according to the prior annotations of the original map to focus on the more complex lane line portions other than the straight line portions.

[0076] In some embodiments, the target portion of the lane line to be tested is segmented according to the original map to obtain a segmented curve of the current frame, including: obtaining a curve equation corresponding to the lane line to be tested, and determining the curvature of multiple target curves; wherein each target curve is a curve composed of a first preset number of adjacent points on the lane line to be tested; calculating the curvature difference between adjacent target curves; when the curvature difference is greater than the preset curvature difference, and multiple target curves correspond to the target portion of the lane line to be tested, segmenting the curve corresponding to the target portion of the lane line to be tested according to the multiple target curves to obtain the segmented curve of the current frame.

[0077] It should be noted that the lane line is composed of many points. In the embodiment of the present disclosure, the lane line to be tested is divided according to the first preset number of adjacent points to obtain multiple target curves. For example, the first preset number is 20. Each target curve obtained by dividing the lane line to be tested is composed of 20 points in units of 20 points, that is, the target curve is a curve composed of 20 adjacent points on the lane line to be tested. Then, the curvature of the multiple target curves is calculated, and then the curvature change is monitored, and the curvature difference between adjacent target curves is calculated, and the curvature difference is compared to see whether it is greater than the preset curvature difference, wherein the preset curvature difference indicates that the curve is the maximum curvature difference of the smooth transition, and if it is greater than the preset curvature difference, it indicates that the adjacent curves are not smoothly transitioned, and if it is less than or equal to the curvature difference, it indicates that the adjacent curves are smoothly transitioned. In the case where the curvature difference between the adjacent target curves is greater than the preset curvature difference, it indicates that the curvature has a sudden change between the adjacent target curves, and it can be considered that the two target curves are not smoothly transitioned, and the two target curves are determined to be partial segmented curves of the current frame. After monitoring the curvature changes of multiple target curves, the segmented curve of the current frame is obtained. The segmented curve of the current frame is obtained by segmenting the cubic curve corresponding to the non-smooth transition area of ​​the lane line to be tested at the current moment.

[0078] For the identification of non-smooth transition areas, a preliminary determination can be made based on the prior annotations in the original map, and then a further determination can be made based on whether the curvature difference between adjacent curves is greater than the preset curvature difference. In the implementation of the present disclosure, the target portion of the lane line to be tested is determined based on the prior annotations in the original map to preliminarily locate the non-smooth transition area; the curvature difference of the lane line to be tested is calculated for segmentation and then compared with the preset curvature difference, which is to locate the non-smooth transition area in a quantitative manner. The combination of the two can accurately locate the non-smooth past area of ​​the lane line to be tested. It can be understood that the segmented curve of the current frame is a curve belonging to the target portion of the lane line to be tested, and the curvature difference between adjacent segmented curves is greater than the preset curvature difference.

[0079] In some embodiments, after obtaining the target portion of the lane line to be measured based on the marking of the lane line in the original map, these target portions are divided according to a preset number of adjacent points to obtain multiple target curves, which can be understood as the target curve corresponding to the non-straight lane; then the curvature difference between the target curves is calculated, and when the curvature difference between the target curves is greater than the preset curvature difference, the segmented curve of the current frame is obtained. It can be understood that if the target curve is a part of the curve corresponding to the target portion of the lane line to be measured, and the curvature difference between adjacent target curves is greater than the preset curvature difference, the segmented curve of the current frame is obtained based on these multiple target curves. The segmented curve is also represented by a feature vector.

[0080] Exemplarily, the curve corresponding to the target part of the lane line to be tested is divided into 50 target curves. For these 50 target curves, the curvature difference between adjacent target curves is calculated. If the curvature difference between these 50 target curves is greater than the preset curvature difference, it can be determined that the target part of the lane line to be tested is a large curvature curve and needs to be paid special attention to. Therefore, these 50 target curves are used as the segmented curves of the current frame.

[0081] The disclosed embodiment focuses on the target part of the lane line, such as ramps, large curvature curves and other non-smooth transition parts, and segments the target part of the lane line to be tested, so as to determine the degree of correlation between each segmented curve and the historical lane line in subsequent steps. In this way, the complex lane lines are focused, which is conducive to building a more accurate vector map.

[0082] S103: Calculate the correlation parameters between the segmented curve and the historical lane line.

[0083] The association parameter is used to indicate the possibility of the segmented curve being associated with the historical lane line. The larger the association parameter, the higher the possibility of the segmented curve being associated with the historical lane line, and the greater the possibility that the segmented curve and the historical lane line belong to the same lane line.

[0084] In some embodiments, after executing step S102 and before executing step S103, it also includes: obtaining historical lane lines and projecting the historical lane lines to the vehicle coordinate system of the current frame; calculating the association parameters between the segmented curve and the historical lane lines includes: calculating the association parameters between the segmented curve and the historical lane lines in the vehicle coordinate system of the current frame.

[0085] Specifically, before calculating the associated parameters between the segmented curve and the historical lane line, the historical lane line is first obtained. The historical lane line is the lane line obtained at a historical moment based on the method provided by the embodiment of the present disclosure. It is the lane line in the vector map obtained after executing steps S101 to S104. It can be the lane line corresponding to the previous frame of the current frame. It should be noted that during the driving of the vehicle, the vehicle coordinate system with the vehicle as the coordinate origin changes. When calculating the associated parameters between the segmented curve and the historical lane line, it is necessary to project the historical lane line to the vehicle coordinate system of the current frame for calculation. The embodiment of the present disclosure provides an implementation method. First, the odometer information is obtained, and the posture change information of the vehicle is determined according to the odometer information. Then, according to the posture change information, the historical frame vehicle coordinate system corresponding to the historical lane line is converted into the vehicle coordinate system of the current frame, so as to project the historical lane line to the vehicle coordinate system of the current frame. The vehicle coordinate system of the current frame includes the historical lane line and the segmented curve of the current frame.

[0086] Furthermore, in the vehicle coordinate system of the current frame, the association parameters between the segmented curve and the historical lane line are calculated.

[0087] On the basis of the above embodiment, the association parameters between the segmented curve and the historical lane line are calculated, including: extracting a second preset number of visual points from the segmented curve; calculating the residual value between each visual point and the historical lane line; calculating the residual mean according to the residual value of each visual point and the second preset number; and taking the inverse of the residual mean as the association parameter.

[0088] Specifically, a second preset number of visual points is extracted from the segmented curve, which may be 20. The residual value between each visual point and the historical lane line is then calculated to obtain the second preset number of residual values, which are accumulated to obtain the deviation value, which is divided by the second preset number to obtain the residual mean of the segmented curve, and the inverse of the residual mean is further used as an associated parameter.

[0089] In the above embodiment, the correlation parameters between each segmented curve and the historical lane line are calculated respectively, so as to determine the correlation degree between the segmented curve and the historical lane line.

[0090] S104. When the association parameter is greater than or equal to a preset association threshold, the segmented curve is associated with the historical lane line to obtain a vector map.

[0091] In some embodiments, when the association parameter is greater than or equal to a preset association threshold, the Hungarian algorithm is used to associate the segmented curve with the largest association parameter with the historical lane line, and specifically, the point cloud corresponding to the segmented curve is stored corresponding to the point cloud of the historical lane line.

[0092] In some embodiments, when the association parameter is greater than or equal to a preset association threshold, the segmented curve is associated with the historical lane line to obtain an appropriate map, including: when the association parameter is greater than or equal to the preset association threshold, determining the horizontal coordinate value of the visual point on the segmented curve, the horizontal coordinate value is used to represent the distance between the visual point and the vehicle; setting the weight coefficient of the visual point according to the horizontal coordinate value, the weight coefficient is inversely proportional to the horizontal coordinate value; fitting is performed according to the historical lane line, the segmented curve and the weight coefficient to obtain a vector map.

[0093] Specifically, when the association parameter is greater than or equal to the preset association threshold, the horizontal coordinate value of the visual point on the segmented curve is determined, and the horizontal coordinate value represents the distance between the visual point and the vehicle. The smaller the horizontal coordinate value, the closer the visual point on the segmented curve is to the vehicle. Then, the weight coefficient of the visual point is set according to the horizontal coordinate value. The weight coefficient is inversely proportional to the horizontal coordinate value. The smaller the horizontal coordinate value, the larger the weight coefficient, indicating that the closer the visual point to the vehicle is, the higher the attention it pays to it. It can be understood that due to factors such as distortion of the fisheye camera, environmental occlusion, and light changes, the identified lane line may be inaccurate, that is, there is an error in the segmented curve, but the reliability of the part closer to the vehicle is higher. The present disclosure sets the weight coefficient of the visual point according to the distance between the visual point and the vehicle. Further, fitting is performed according to the historical lane line, the segmented curve and the weight coefficient to obtain a more accurate vector map. Optionally, the least squares fitting is performed on the historical lane line, the segmented curve and the weight coefficient to obtain the cubic curve equation of the same lane line.

[0094] Based on the above embodiment, fitting is performed according to historical lane lines, segmented curves and weight coefficients to obtain a vector map, including: fitting according to historical lane lines, segmented curves and weight coefficients, filtering the fitting results to remove noise in the fitting results, and obtaining a vector map.

[0095] Specifically, after fitting according to the historical lane lines, segmented curves and weight coefficients to obtain the fitting results (the cubic curve equation of the same lane line), the fitting results are filtered (such as Kalman filtering) to remove the noise in the fitting results and realize the fusion of the historical lane lines and segmented curves, so as to associate the lane lines to be detected in the current frame with the historical lane lines to obtain a vector map.

[0096] The above embodiment takes into account the noise that may exist after the historical lane lines and the segmented curve fitting, and uses Kalman filtering to perform denoising to fuse and obtain more accurate lane lines.

[0097] In some embodiments, the above operation is performed on multiple consecutive frames after the current frame to obtain the characteristic vector of the associated lane line, and generate the cubic curve equation of the lane line to update the lane line and the vector map in real time; by processing the subsequent bird's-eye view, the consistency of the lane line between consecutive frames is ensured, so that the generated vector map is more stable and better applied to vehicle driving control.

[0098] In some embodiments, after obtaining the cubic curve equation of the lane line, resampling is performed to obtain a point column corresponding to the lane line to obtain a vector map, and the vector map includes a point column corresponding to at least one lane line. After obtaining the cubic curve equation of the same lane line in adjacent frames, resampling is performed to improve the accuracy of the lane line.

[0099] In summary, the present disclosure provides a method for constructing a vector map, which first obtains the lane line to be tested identified based on the bird's-eye view of the current frame, and the original map with the target part of the lane line marked, and then segments the target part of the lane line to be tested according to the original map to obtain the segmented curve of the current frame, and further calculates the association parameter between the segmented curve and the historical lane line, and the association parameter is used to indicate the possibility of the segmented curve being associated with the historical lane line; when the association parameter is greater than or equal to a preset association threshold, the segmented curve is associated with the historical lane line, thereby obtaining a vector map. The present disclosure segments the target part of the lane line to be tested corresponding to the current frame, and separately determines the possibility of each segmented curve being associated with the historical lane line, thereby associating the segmented curve with a high possibility with the historical lane line, accurately identifying the same lane line in the current frame and the historical frame, and improving the consistency and stability of the lane lines in the vector map.

[0100] like Figure 2 As shown, Figure 2 A schematic diagram of a vector map construction device provided in an embodiment of the present disclosure, the device comprising:

[0101] An acquisition module 201 is used to acquire a lane line to be tested and an original map, wherein the lane line to be tested is identified based on a bird's-eye view of the current frame, and a target portion of the lane line is marked in the original map;

[0102] A segmentation module 202, used to segment the target part of the lane line to be tested according to the original map to obtain a segmentation curve of the current frame;

[0103] A calculation module 203, used to calculate an association parameter between the segmented curve and the historical lane line, where the association parameter is used to indicate the possibility of association between the segmented curve and the historical lane line;

[0104] The association module 204 is used to associate the segmented curve with the historical lane line to obtain a vector map when the association parameter is greater than or equal to a preset association threshold.

[0105] As an optional implementation of the embodiment of the present disclosure, the segmentation module 202 is specifically used to: obtain a curve equation corresponding to the lane line to be tested, and determine the curvatures of multiple target curves; wherein each target curve is a curve composed of a first preset number of adjacent points on the lane line to be tested;

[0106] Calculate the curvature difference between adjacent target curves;

[0107] When the curvature difference is greater than a preset curvature difference and the multiple target curves correspond to the target portion of the lane line to be measured, the target portion of the lane line to be measured is segmented according to the multiple target curves to obtain a segmented curve of the current frame.

[0108] As an optional implementation of the embodiment of the present disclosure, the calculation module 203 is further used to: obtain historical lane lines, and project the historical lane lines to the vehicle coordinate system of the current frame;

[0109] The calculation module 203 is specifically used to calculate the association parameters between the segmented curve and the historical lane line in the vehicle coordinate system of the current frame.

[0110] As an optional implementation of the embodiment of the present disclosure, the calculation module 203 is specifically used to: obtain odometer information, and determine the posture change information of the vehicle according to the odometer information;

[0111] According to the posture change information, the vehicle coordinate system of the historical frame corresponding to the historical lane line is converted into the vehicle coordinate system of the current frame, so as to project the historical lane line into the vehicle coordinate system of the current frame.

[0112] As an optional implementation of the embodiment of the present disclosure, the calculation module 203 is specifically used to: extract a second preset number of visual points from the piecewise curve;

[0113] Calculate the residual value between each visual point and the historical lane line;

[0114] The residual mean is calculated based on the residual value of each visual point and the second preset number;

[0115] The reciprocal of the residual mean is taken as the correlation parameter.

[0116] As an optional implementation of the embodiment of the present disclosure, the association module 204 is specifically used to: determine the abscissa value of the visual point on the segmented curve when the association parameter is greater than or equal to a preset association threshold, where the abscissa value is used to represent the distance between the visual point and the vehicle;

[0117] The weight coefficient of the visual point is set according to the horizontal coordinate value, and the weight coefficient is inversely proportional to the horizontal coordinate value;

[0118] Fitting is performed based on historical lane lines, segmented curves and weight coefficients to obtain a vector map.

[0119] As an optional implementation of the embodiment of the present disclosure, the association module 204 is specifically used to: perform fitting according to historical lane lines, segmented curves and weight coefficients;

[0120] The fitting results are filtered to remove noise in the fitting results and obtain a vector map.

[0121] like Figure 3 As shown, Figure 3 The present disclosure provides a schematic diagram of the structure of an electronic device, which includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, and when the computer program is executed by the processor, each process of the method for constructing a vector map in the above method embodiment is implemented. The same technical effect can be achieved, and it will not be repeated here to avoid repetition.

[0122] The embodiment of the present disclosure provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, each process of the method for constructing a vector map in the above method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0123] The computer readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0124] The embodiment of the present disclosure provides a computer program product, which stores a computer program. When the computer program is executed by a processor, each process of the method for constructing a vector map in the above method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0125] Those skilled in the art will appreciate that the embodiments of the present disclosure may be provided as methods, systems, or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present disclosure may take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code.

[0126] In several embodiments provided by the present disclosure, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart, can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0127] In the present disclosure, the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0128] In the present disclosure, memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0129] In the present disclosure, computer-readable media include permanent and non-permanent, removable and non-removable storage media. Storage media can be implemented by any method or technology to store information, and the information can be computer-readable instructions, data structures, modules of programs or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. According to the definition in this article, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0130] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0131] The above are only specific embodiments of the present disclosure, so that those skilled in the art can understand or implement the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to these embodiments herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for constructing a vector map. It is characterized in that include: Obtaining a lane line to be tested and an original map, wherein the lane line to be tested is identified based on a bird's-eye view of the current frame, and a target portion of the lane line is marked in the original map; Segmenting the target portion of the lane line to be tested according to the original map to obtain a segmented curve of the current frame; Calculating an association parameter between the segmented curve and the historical lane line, wherein the association parameter is used to indicate a possibility of association between the segmented curve and the historical lane line; When the association parameter is greater than or equal to a preset association threshold, the segmented curve is associated with the historical lane line to obtain a vector map.

2. The method according to claim 1, It is characterized in that The step of segmenting the target portion of the lane line to be tested according to the original map to obtain the segmented curve of the current frame includes: Obtaining a curve equation corresponding to the lane line to be tested, and determining the curvatures of a plurality of target curves; wherein each of the target curves is a curve formed by a first preset number of adjacent points on the lane line to be tested; Calculate the curvature difference between adjacent target curves; When the curvature difference is greater than a preset curvature difference and the multiple target curves correspond to the target portion of the lane line to be measured, the target portion of the lane line to be measured is segmented according to the multiple target curves to obtain the segmented curves of the current frame.

3. The method according to claim 1, It is characterized in that After segmenting the target portion of the lane line to be tested according to the original map to obtain the segmented curve of the current frame and before calculating the association parameter between the segmented curve and the historical lane line, the method further includes: Acquire historical lane lines, and project the historical lane lines into the vehicle coordinate system of the current frame; The calculating the association parameter between the segmented curve and the historical lane line includes: In the vehicle coordinate system of the current frame, an association parameter between the segmented curve and the historical lane line is calculated.

4. The method according to claim 3, It is characterized in that The acquiring of historical lane lines and projecting the historical lane lines to the vehicle coordinate system of the current frame includes: Obtain odometer information, and determine the vehicle's posture change information based on the odometer information; According to the posture change information, the vehicle coordinate system of the historical frame corresponding to the historical lane line is converted into the vehicle coordinate system of the current frame, so as to project the historical lane line into the vehicle coordinate system of the current frame.

5. The method according to claim 1 or 3, It is characterized in that The calculating the association parameter between the segmented curve and the historical lane line includes: Extracting a second preset number of visual points from the segmented curve; Calculating a residual value between each of the visual points and the historical lane line; Calculating a residual mean value according to the residual value of each visual point and the second preset number; The reciprocal of the residual mean is used as the correlation parameter.

6. The method according to claim 1, It is characterized in that When the association parameter is greater than or equal to a preset association threshold, the segmented curve is associated with the historical lane line to obtain a vector map, including: When the association parameter is greater than or equal to a preset association threshold, determining a horizontal coordinate value of the visual point on the segmented curve, wherein the horizontal coordinate value is used to represent the distance between the visual point and the vehicle; Setting a weight coefficient of the visual point according to the horizontal coordinate value, wherein the weight coefficient is in inverse proportion to the horizontal coordinate value; Fitting is performed according to the historical lane lines, the segmented curves and the weight coefficients to obtain the vector map.

7. The method according to claim 6, It is characterized in that The fitting according to the historical lane line, the segmented curve and the weight coefficient to obtain the vector map includes: Performing fitting according to the historical lane line, the segmented curve and the weight coefficient; The fitting result is filtered to remove noise in the fitting result to obtain the vector map.

8. A device for constructing a vector map, It is characterized in that include: An acquisition module, used to acquire a lane line to be tested and an original map, wherein the lane line to be tested is identified based on a bird's-eye view of the current frame, and a target portion of the lane line is marked in the original map; A segmentation module, used for segmenting the target part of the lane line to be tested according to the original map to obtain the segmentation curve of the current frame; A calculation module, used for calculating an association parameter between the segmented curve and the historical lane line, wherein the association parameter is used for indicating a possibility of association between the segmented curve and the historical lane line; The association module is used to associate the segmented curve with the historical lane line to obtain a vector map when the association parameter is greater than or equal to a preset association threshold.

9. An electronic device, It is characterized in that include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the method for constructing a vector map as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, It is characterized in that include: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for constructing a vector map according to any one of claims 1 to 7 is implemented.

11. A vehicle, It is characterized in that It comprises the vector map construction device as claimed in claim 8, or the electronic device as claimed in claim 9.

Images