Lightweight map making method and system, vehicle and medium

Abstract

The invention relates to the technical field of intelligent driving, and discloses a lightweight map making method and system, a vehicle and a medium, and the method comprises the steps: obtaining a driving track, road information and navigation information of a target vehicle; based on the navigation information and the road information, carrying out fusion processing on the driving track to obtain a target track; extracting a target scene from the road information, and constructing a scene data set; and making a map of the target vehicle based on the target trajectory and the scene data set. According to the method, the driving track and the road condition in the vehicle driving process are acquired, track fusion and scene information extraction of key road conditions such as lane number change, intersections and stop lines are performed, and the map is made based on the fused track and scene information, so that the data volume required for map making can be greatly reduced, the map making precision and efficiency are improved, and the map making cost is reduced. The updating period is shortened, and the intelligent navigation requirement of the vehicle is met.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent driving, in particular to a lightweight map making method and system, vehicle and medium. BACKGROUND

[0002] At present, with the rapid development of deep learning technology, artificial intelligence is applied to many fields and achieves good application effect, and intelligent driving is a popular field and attracts much attention.

[0003] In recent years, intelligent driving technology greatly facilitates people's life from emergency assistance function to partial automatic driving function, and releases the driver from long-time repeated simple actions, and avoids the resulting mistakes. However, in the process of rapid development of intelligent driving, new problems arise, that is, with the wide application of high-precision maps, the produced maps have the problems of large data volume, long update cycle, low precision, and are difficult to meet the intelligent navigation requirements of the vehicle. SUMMARY

[0004] Therefore, the present application provides a lightweight map making method and system, vehicle and medium to solve the problems of large data volume, long update cycle, low map precision, and difficulty in meeting the intelligent navigation requirements of the vehicle.

[0005] In a first aspect, the present application provides a lightweight map making method, which comprises:

[0006] obtaining the driving trajectory, road information and navigation information of the target vehicle;

[0007] fusing the driving trajectory based on the navigation information and the road information to obtain a target trajectory;

[0008] extracting a target scene from the road information to construct a scene data set;

[0009] making a map of the target vehicle based on the target trajectory and the scene data set.

[0010] The present application fuses the driving trajectory and the road condition obtained during the driving process of the vehicle, extracts the scene of key road conditions such as lane number change, intersection and stop line, and makes a map based on the fused trajectory and scene information, which can greatly reduce the data volume required for making a map, greatly improve the precision and efficiency of map making, reduce the map update cycle, and meet the intelligent navigation requirements of the vehicle.

[0011] In an optional embodiment, the fusing of the driving trajectory based on the navigation information and the road information to obtain a target trajectory comprises:

[0012] The road structure based on the navigation information is used to segment road information, to obtain a plurality of road segments and to identify the road segments respectively;

[0013] For each road segment identification, a plurality of same-identification road segments are obtained by screening driving tracks with the same road segment identification respectively;

[0014] It is determined whether the number of tracks in each same-identification road segment is less than a first preset track threshold value respectively;

[0015] When the number of tracks is not less than the first preset track threshold value, the driving tracks in the corresponding same-identification road segment are fused to obtain a target track;

[0016] When the number of tracks is less than the first preset track threshold value, the steps of obtaining the driving track of the target vehicle, the road information and the navigation information are returned.

[0017] The present application segments road information based on road structure, and fuses the driving tracks of vehicles corresponding to the same road segment, which can reduce the amount of data required in the map making process, and help to speed up the map making process, and generate a lightweight vehicle map.

[0018] In an optional embodiment, the driving tracks in the corresponding same-identification road segment are fused to obtain a target track, including:

[0019] The driving tracks in each same-identification road segment are mapped to the same coordinate system;

[0020] The fitting degree of all driving tracks in each same-identification road segment is calculated, and the fitting tracks are obtained based on the fitting degree calculation result;

[0021] The fitting tracks corresponding to all same-identification road segments are integrated to obtain a target track.

[0022] The present application calculates the fitting degree of the driving track, and fuses the track based on the fitting degree calculation result, which can greatly reduce the amount of data required for making a map, and improve the efficiency of map making.

[0023] In an optional embodiment, the fitting degree of all driving tracks in each same-identification road segment is calculated, and the fitting tracks are obtained based on the fitting degree calculation result, including:

[0024] The lateral distance between two adjacent driving tracks is calculated for all driving tracks in each same-identification road segment respectively;

[0025] The fitting degree calculation result of the two adjacent driving tracks is determined based on the lateral distance;

[0026] screening all driving trajectories with a fitting degree calculation result greater than a preset fitting degree threshold value, and constructing a trajectory dataset;

[0027] judging whether the number of trajectories in the trajectory dataset is less than a second preset trajectory threshold value;

[0028] when the number of trajectories is not less than the second preset trajectory threshold value, performing trajectory fusion on the trajectory dataset, and correspondingly obtaining a fused trajectory;

[0029] when the number of trajectories is less than the second preset trajectory threshold value, returning to the step of obtaining the driving trajectory, the road information and the navigation information of the target vehicle.

[0030] The present application determines the fitting degree calculation result by the lateral distance of two adjacent driving trajectories, and designs a double determination process of the fitting degree calculation result and the preset fitting degree threshold value, and the number of trajectories in the trajectory dataset and the corresponding preset trajectory threshold value for trajectory fusion, which not only reduces the data amount of the driving trajectory, but also guarantees the acquisition accuracy of the fused trajectory, and helps to improve the accuracy and efficiency of subsequent map making.

[0031] In an optional embodiment, a target scene is extracted from the road information to construct a scene dataset, including:

[0032] screening the road information corresponding to the target trajectory;

[0033] extracting the target scene from the road information based on a preset scene element, and the preset scene element includes a lane line, a building and a road structure;

[0034] integrating all target scenes to obtain the scene dataset.

[0035] The present application considers the extraction of key scene information such as lane lines, buildings and road structures during vehicle driving, and can obtain rich road information to guarantee the accuracy of map making.

[0036] In an optional embodiment, a map of the target vehicle is made based on the target trajectory and the scene dataset, including:

[0037] associating all target scenes in the scene dataset with the target trajectory respectively to obtain an associated trajectory of the target vehicle;

[0038] generating a map of the target vehicle based on the associated trajectory.

[0039] The present application associates the target scene with the target trajectory and generates a corresponding map, which can greatly reduce the amount of data required for making a map, improve the accuracy and efficiency of map making, reduce the map update cycle, and greatly meet the intelligent navigation needs of vehicles.

[0040] In an optional embodiment, before making the map of the target vehicle based on the target trajectory and the scene data set, the lightweight map making method further comprises:

[0041] generating a standard definition map based on the navigation information;

[0042] respectively judging whether the target scene in the scene data set matches the corresponding position in the standard definition map;

[0043] when the target scene matches the corresponding position in the standard definition map, retaining the corresponding target scene;

[0044] when the target scene does not match the corresponding position in the standard definition map, deleting the corresponding target scene and updating the scene data set.

[0045] The present application matches the target scene with the corresponding generated standard definition map, which can guarantee the authenticity and reliability of the target scene and help improve the accuracy of the made map.

[0046] In a second aspect, the present application provides a lightweight map making system, which comprises:

[0047] an acquisition module for acquiring the driving trajectory, road information and navigation information of the target vehicle;

[0048] a fusion module for performing fusion processing on the driving trajectory based on the navigation information and the road information to obtain the target trajectory;

[0049] an extraction module for extracting the target scene from the road information to construct a scene data set;

[0050] a making module for making the map of the target vehicle based on the target trajectory and the scene data set.

[0051] The lightweight map making system of the present application can greatly reduce the amount of data required for making the map by performing trajectory fusion and scene extraction on the driving trajectory and road conditions during the driving process of the vehicle and making the map based on the fused trajectory and scene information, thereby improving the accuracy and efficiency of map making and meeting the intelligent navigation requirements of the vehicle.

[0052] In a third aspect, the present application provides a vehicle, which comprises a controller, the controller comprising a memory and a processor, the memory and the processor being communicatively connected with each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the lightweight map making method of the first aspect or any of the corresponding embodiments thereof.

[0053] In a fourth aspect, the present application provides a computer readable storage medium, which stores computer instructions for causing a computer to execute the lightweight map making method of the first aspect or any of the corresponding embodiments thereof.

[0054] The lightweight map making method of the present application records the vehicle driving track by using the vehicle positioning system and collects the road conditions during the vehicle driving by using the vehicle camera, and the fusion processing of the driving track can greatly reduce the data amount required for making the map, the collected road conditions are analyzed and the key scenes such as the lane number change, the intersection and the stop line are extracted therefrom, the fusion track and the key scenes are processed to generate the corresponding lightweight map, and the lightweight map has the advantages of small data amount, short update cycle, high and stable map precision, and greatly meets the intelligent navigation requirements of the vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0055] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the drawings required to be used in the specific embodiments or the prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0056] Figure 1 is a flowchart of the lightweight map making method according to the embodiment of the present application;

[0057] Figure 2 is a flowchart of another lightweight map making method according to the embodiment of the present application;

[0058] Figure 3 is a flowchart of still another lightweight map making method according to the embodiment of the present application;

[0059] Figure 4 is a schematic diagram of the vehicle equipment for data collection;

[0060] Figure 5 is a schematic diagram of the data segmentation processing;

[0061] Figure 6 is a schematic diagram of the track fusion;

[0062] Figure 7 is a schematic diagram of the node scene;

[0063] Figure 8 is a schematic diagram of the effect of generating the map;

[0064] Figure 9 is a structural block diagram of the lightweight map making system according to the embodiment of the present application;

[0065] Figure 10 is a structural schematic diagram of a controller of a vehicle of an embodiment of the present application. DETAILED DESCRIPTION

[0066] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0067] The embodiments of the present application provide a lightweight map making method, and it should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that here.

[0068] In the present embodiment, a lightweight map making method is provided, which is applied to a controller in a new energy vehicle, such as a single-chip microcomputer, a microprocessor, etc. Figure 1 is a flowchart of the lightweight map making method according to the embodiments of the present application, as shown in Figure 1 The flowchart includes the following steps:

[0069] Step S101, obtaining the driving track, road information, and navigation information of the target vehicle.

[0070] It should be noted that the obtaining method of the driving track, road information, and navigation information of the target vehicle in the present embodiment is not limited here, and is adaptively selected based on actual needs. For example, the driving track of the target vehicle is recorded by using a vehicle-mounted recorder or a vehicle-mounted positioning system installed on the target vehicle; the road information (i.e. image data) around the vehicle is collected by a vehicle-mounted camera; and a corresponding navigation route is generated in response to the vehicle navigation information set by the user and is displayed through the vehicle; the above content is only used as an example and is not limited thereto.

[0071] In actual application, the driving track, road information, and navigation information of the target vehicle are stored in the corresponding vehicle end or cloud end, and the specific storage method is not limited here and is adaptively adjusted based on actual needs. For example, the storage can be set to mark the data with a data source label, and the data source label is in the form of storage path + data file name, wherein the data file name includes data collection time, vehicle frame number, etc. Specifically, through the storage and marking of the vehicle data, the data collection time, vehicle information, etc. can be traced, and the original data can be obtained to troubleshoot corresponding problems when the vehicle fails.

[0072] Step S102, fusion processing is performed on the driving trajectories based on the navigation information and the road information to obtain a target trajectory.

[0073] It should be noted that the fusion of the driving trajectories in this embodiment aims to fuse driving trajectories with high similarity to reduce the data required for map making and speed up the map generation process. The specific fusion means in this embodiment are not limited here and can be adaptively adjusted based on actual requirements. For example, a plurality of driving trajectories are fitted by a curve fitting algorithm to obtain a target trajectory, wherein the curve fitting algorithm includes linear regression, polynomial fitting, least squares method, etc., which are only illustrative and not limited thereto, and can be adaptively adjusted based on actual requirements.

[0074] Step S103, a target scene is extracted from the road information to construct a scene data set.

[0075] It should be noted that the type and content of the target scene in this embodiment are not limited here and can be adaptively adjusted based on the scenes that affect vehicle driving during actual vehicle driving, such as changes in the number of lanes, divergences, mergings, tunnels, toll stations, intersections, roundabouts, construction barriers, etc. These are only illustrative.

[0076] In a specific embodiment, for the target vehicle's current lane, if the right ground lane line changes from a solid line to a dashed line, it indicates that the vehicle can change lanes to the right under the condition of ensuring safety with lane-changing conditions, so the position of the lane line change can also be used as a target scene.

[0077] Step S104, a map of the target vehicle is made based on the target trajectory and the scene data set.

[0078] The lightweight map making method of the embodiment of the present application fuses the driving trajectories and the road conditions obtained during the driving of the vehicle, extracts scenes such as changes in the number of lanes, intersections, stop lines, etc., and makes a map based on the fused trajectories and scene information, which greatly reduces the amount of data required for making a map, improves the accuracy and efficiency of map making, reduces the map update cycle, and can meet the intelligent navigation needs of vehicles.

[0079] In this embodiment, a lightweight map making method is provided, Figure 2 is a flowchart of another lightweight map making method according to an embodiment of the present application, as shown in Figure 2 The flowchart includes the following steps:

[0080] Step S201, the driving trajectory, road information and navigation information of a target vehicle are obtained. For details, see step S101 of the embodiment shown in Figure 1 , which will not be repeated here.

[0081] In step S202, the driving track is fused based on the navigation information and the road information to obtain a target track.

[0082] Specifically, the step S202 includes the following steps.

[0083] In step S2021, the road information is segmented based on the road structure of the navigation information to obtain a plurality of road segments and to identify the road segments respectively.

[0084] It should be noted that the road structure in the embodiment is referred to as a link, and the link refers to a road segment, which is a basic unit of a road model in a navigation system. In actual applications, since different routes can pass through the same road segment, the same road segment in different routes collected at different times can be merged together for processing after data segmentation. The link in the navigation is selected as the data segmentation criterion based on the road structure changes, such as the changes in the intersection and the number of lanes, which is helpful for the effective fusion of subsequent tracks and accelerates the map making process. Specifically, the navigation link is used as the data segmentation basis, which not only has the advantages of high reliability and small amount of calculation, but also can avoid the calculation errors caused by the fact that the road image obtained by the camera does not collect the entire road surface.

[0085] In the embodiment, the specific way of identifying the plurality of road segments is not limited here, and the corresponding data label can be set based on actual needs to mark the data source and the road name. For example, the a route collected by the camera of the vehicle includes a plurality of road segments, which are marked as a1, a2, a3, etc. respectively; the b route collected by the camera of the vehicle includes a plurality of road segments, which are marked as a2, a3, b1, b2, b3, etc. respectively, which are only exemplary.

[0086] In step S2022, for each road segment identification, the driving track with the same road segment identification is selected to obtain a plurality of road segments with the same identification.

[0087] In the embodiment, the plurality of driving tracks corresponding to the same road segment identification are selected to obtain a plurality of track data sets corresponding to the same road segment identification.

[0088] In step S2023, it is judged whether the number of tracks in each road segment with the same identification is less than a first preset track threshold.

[0089] In the embodiment, the specific value of the first preset track threshold is not limited here and is adaptively selected based on actual needs. For example, the first preset track threshold is 5, which is only exemplary.

[0090] In step S2024, when the number of trajectories is not less than the first preset trajectory threshold, the trajectories corresponding to the same identified road segments are fused to obtain target trajectories.

[0091] Specifically, in step S2024, the trajectories corresponding to the same identified road segments are fused to obtain target trajectories, including:

[0092] In step A1, the trajectories in each same identified road segment are mapped to the same coordinate system.

[0093] In this embodiment, the specific type of the coordinate system is not limited here and is adaptively adjusted based on actual needs. For example, the coordinate system is a vehicle coordinate system, which is only illustrative.

[0094] In step A2, the fitting degrees of all trajectories in each same identified road segment are calculated, and the trajectories are fused based on the fitting degree calculation results to obtain fused trajectories.

[0095] Specifically, in step A2, the following steps are included:

[0096] In step A21, for all trajectories in each same identified road segment, the lateral distance between two adjacent trajectories is calculated.

[0097] In actual applications, since the trajectories of the vehicle are curved, the lateral distance between the two adjacent trajectories in this embodiment can be determined by sampling the same positions on the two corresponding trajectories and calculating the distance between the two sampling points, and the lateral distance between the two adjacent trajectories is determined by multiple sampling points. It should be noted that the way of determining the lateral distance between the two adjacent trajectories by multiple sampling points in this embodiment is not limited here and is adaptively adjusted based on the actual project precision requirements. For example, the average value of all sampling points is calculated as the lateral distance; or the 90th percentile of all sampling points is taken as the lateral distance, which is only illustrative and is not limited thereto.

[0098] In step A22, the fitting degree calculation result of the two adjacent trajectories is determined based on the lateral distance.

[0099] In this embodiment, the smaller the lateral distance, the more similar the two adjacent trajectories are, and the greater the corresponding fitting degree calculation result.

[0100] In step A23, all trajectories with fitting degree calculation results greater than a preset fitting degree threshold are screened out to construct a trajectory dataset.

[0101] In this embodiment, the specific value of the preset fitting degree threshold is not limited here and is adaptively set according to actual needs.

[0102] Step A24, judging whether the number of trajectories in the trajectory dataset is less than a second preset trajectory threshold.

[0103] In the embodiment, the specific value of the second preset trajectory threshold is not limited here, and is adaptively selected based on actual requirements. For example, the second preset trajectory threshold is 4, which is only used as an example.

[0104] Step A25, when the number of trajectories is not less than the second preset trajectory threshold, performing trajectory fusion on the trajectory dataset to obtain a fused trajectory.

[0105] It should be noted that the trajectory fusion in the embodiment is to fuse all trajectories in the trajectory dataset into one trajectory.

[0106] Step A26, when the number of trajectories is less than the second preset trajectory threshold, returning to step S201 of acquiring the driving trajectory, the road information and the navigation information of the target vehicle.

[0107] The embodiment of the application determines the fitting degree calculation result by the lateral distance of the two adjacent driving trajectories, and designs a double determination process of the fitting degree calculation result and the preset fitting degree threshold, and the number of trajectories in the trajectory dataset and the corresponding preset trajectory threshold for trajectory fusion, which not only reduces the data amount of the driving trajectory, but also guarantees the acquisition accuracy of the fused trajectory, and helps to improve the accuracy and efficiency of subsequent map making.

[0108] Step A3, integrating the fused trajectories corresponding to all the same identified road segments to obtain a target trajectory.

[0109] In actual application, a complete route includes multiple different road segments. In the embodiment, the fused trajectories corresponding to all the same identified road segments are spliced to obtain a corresponding target trajectory. Specifically, by performing fitting degree calculation on the driving trajectory and performing trajectory fusion based on the fitting degree calculation result, the data amount required for making a map can be greatly reduced, and the efficiency of map making can be improved.

[0110] Step S2025, when the number of trajectories is less than the first preset trajectory threshold, returning to step S201 of acquiring the driving trajectory, the road information and the navigation information of the target vehicle.

[0111] The embodiment of the application segments the road information based on the road structure, and fuses the vehicle driving trajectories corresponding to the same road segment, which can reduce the data amount required in the map making process, and helps to speed up the map making process and generate a lightweight vehicle map.

[0112] Step S203, extracting a target scene from the road information to construct a scene dataset.

[0113] Specifically, the above step S203 includes:

[0114] Step S2031, screening road information corresponding to the target trajectory.

[0115] In this embodiment, the road information corresponding to the target trajectory is screened through the data source label, i.e. the image data recorded by the vehicle-mounted camera corresponding to the target trajectory.

[0116] Step S2032, extracting a target scene from the road information based on a preset scene element, the preset scene element including a lane line, a building and a road structure.

[0117] It should be noted that the target scene extracted in this embodiment includes a scene type and a coordinate point, wherein the scene type is any one of the preset scene elements. Specifically, a large amount of scene-related image data such as lane lines, buildings and road structures can be collected, and a preset deep learning model can be used to learn the image data and screen out target scenes that meet the definition of special scenes.

[0118] Step S2033, integrating all target scenes to obtain a scene dataset.

[0119] In this embodiment, different types of scenes are classified and stored, and a scene dataset is obtained based on all scene data. Specifically, considering the extraction of key scene information such as lane lines, buildings and road structures during vehicle driving, rich road information can be obtained to ensure the accuracy of map making.

[0120] Step S204, making a map of the target vehicle based on the target trajectory and the scene dataset.

[0121] Specifically, the above step S204 includes:

[0122] Step S2041, associating all target scenes in the scene dataset with the target trajectory respectively to obtain an associated trajectory of the target vehicle.

[0123] Step S2042, generating a map of the target vehicle based on the associated trajectory.

[0124] The embodiment of the application associates the target scene with the target trajectory and generates a corresponding map, which can greatly reduce the amount of data required for making a map, improve the accuracy and efficiency of map making, reduce the map update cycle, and greatly meet the intelligent navigation needs of vehicles.

[0125] It should be noted that before making a map of the target vehicle based on the target trajectory and the scene dataset, the lightweight map making method of the embodiment also considers the reliability of extracting the target scene and designs a corresponding target scene verification process, which specifically includes:

[0126] Step B1, generating a standard definition map based on navigation information.

[0127] It should be noted that the standard definition map (Standard Definition map, SD) is a kind of car machine map, which is used for navigation function.

[0128] Step B2, respectively judging whether the target scene in the scene data set and the corresponding position in the standard definition map match.

[0129] In the embodiment, the matching means of the target scene and the corresponding position in the standard definition map is not limited here, which is set based on actual demand. For example, the coordinate points of the target scene and the coordinate points of the corresponding position in the standard definition map are subtracted, and the matching result is determined based on the size relationship between the difference value and the preset difference value threshold.

[0130] Step B3, when the target scene and the corresponding position in the standard definition map match, the corresponding target scene is retained.

[0131] In the embodiment, if the difference value between the coordinate points of the target scene and the coordinate points of the corresponding position in the standard definition map is less than the preset difference value threshold, it is determined that the two match, and the target scene is real and reliable, which can be retained.

[0132] Step B4, when the target scene and the corresponding position in the standard definition map do not match, the corresponding target scene is deleted, and the scene data set is updated.

[0133] In the embodiment, if the difference value between the coordinate points of the target scene and the coordinate points of the corresponding position in the standard definition map is not less than the preset difference value threshold, it is determined that the two do not match, and the target scene is unreliable, which needs to be deleted and the scene data set is updated. Specifically, the target scene is matched with the corresponding generated standard definition map, which can guarantee the authenticity and reliability of the target scene, and is helpful to improve the accuracy of the produced map.

[0134] In a specific embodiment, Figure 3 is a flowchart of another lightweight map production method according to an embodiment of the application, as shown in Figure 3 , the flow includes the following steps:

[0135] Step C1, data acquisition.

[0136] In the embodiment, the corresponding data acquisition is performed by the vehicle-mounted device for data acquisition of the target vehicle. Specifically, the data acquisition vehicle-mounted device is as shown in Figure 4 , which includes Figure 4It is known that it is composed of a car, a set of vehicle positioning, four cameras (front camera, panoramic camera and rear camera), and the collected data is uploaded to the cloud server through data recording and uploading software. Specifically, the vehicle positioning records the vehicle trajectory, the four vehicle cameras record the road information, and the navigation information is recorded and uploaded to the cloud server.

[0137] Step C2, data segmentation.

[0138] In this embodiment, the recorded collected data is segmented and stored in combination with the navigation information. Specifically, the collected data uploaded to the cloud server is segmented and processed, and the navigation information recorded in step C1 is needed for processing. The link in the navigation information is used as the basis for data segmentation processing, and the segmented data is stored, Figure 5 is a schematic diagram of data segmentation processing. It should be noted that the same road segment data is labeled and stored in the same path.

[0139] Step C3, trajectory fusion.

[0140] In this embodiment, when the number of collected data of the same road segment reaches five times in step C2, the five trajectories are aggregated. First, the fitting degree of the five trajectories is checked, and the trajectories with too large fitting degree are excluded. If the excluded trajectories exceed one, the processing is terminated, and the trajectories are reprocessed when new data is uploaded to the road segment and the fitting degree of the trajectories reaches four within the threshold value. The four trajectories with fitting degree within the threshold value are fitted and fused into one trajectory, and the data source label is marked.

[0141] In a specific embodiment, when the number of collected data of the same road segment reaches five times, i.e. the number of trajectories corresponding to the first processing of the road segment is five, the trajectory fusion condition is met, and the five trajectories are fused. Specifically, first, the fitting degree of the five trajectories is checked, and the trajectories with too large fitting degree are excluded. If the excluded trajectories exceed one, the processing is terminated, and the excluded data is deleted. When new trajectories are input and the number of trajectories reaches five again, the fitting degree is checked again, i.e. whether the number of trajectories with fitting degree within the threshold value reaches the corresponding threshold value. If the number of data with fitting degree within the threshold value is four or more, the data is fitted and fused into one trajectory, Figure 6 is a schematic diagram of trajectory fusion.

[0142] Step C4, node extraction.

[0143] In the embodiment, for the track successfully fused in step C3, the recorded image corresponding to the data source label of the track is processed to extract the lane number change, divergence, confluence, tunnel, toll station, intersection, roundabout, construction fence and other special scenes, and the scene positioning coordinates and records are extracted from the track, Figure 7 is a schematic diagram of a node scene.

[0144] Step C5, node verification.

[0145] In the embodiment, the node scene extracted in step C4 is compared and analyzed with the SD map to confirm whether the scene type and position are correct and whether there is omission.

[0146] Step C6, track node management.

[0147] In the embodiment, the node scene verified through step C5 is stored in association with the track generated in step C3.

[0148] Step C7, data mapping.

[0149] In the embodiment, the data stored in step C6 is mapped, Figure 8 is an effect schematic diagram of a generated map.

[0150] In summary, the lightweight map making method of the embodiment can greatly reduce the amount of data required for making a map, improve the accuracy and efficiency of map making, reduce the update cycle, and meet the intelligent navigation needs of vehicles.

[0151] In the embodiment, a lightweight map making system is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and will not be described again. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the system described in the following embodiments is preferably implemented in software, hardware or a combination of software and hardware is also possible and is contemplated.

[0152] The present application provides a lightweight map making system, as shown in Figure 9 The system comprises:

[0153] The acquisition module 901 is configured to acquire the driving track, road information and navigation information of the target vehicle.

[0154] The fusion module 902 is configured to perform fusion processing on the driving track based on the navigation information and the road information to obtain a target track.

[0155] The extraction module 903 is configured to extract a target scene from the road information, and construct a scene dataset.

[0156] The production module 904 is configured to produce a map of the target vehicle based on the target trajectory and the scene dataset.

[0157] In some optional embodiments, the fusion module 902 includes a first fusion sub-module, a second fusion sub-module, a third fusion sub-module, a fourth fusion sub-module, and a fifth fusion sub-module. The first fusion sub-module is configured to perform data segmentation on the road information based on a road structure of the navigation information, to obtain a plurality of road segments and perform road segment identification on the road segments respectively. The second fusion sub-module is configured to filter, for each road segment identification, a driving trajectory with the same road segment identification to obtain a plurality of same-identification road segments. The third fusion sub-module is configured to determine whether a number of trajectories in each same-identification road segment is less than a first preset trajectory threshold. The fourth fusion sub-module is configured to perform trajectory fusion on the driving trajectories in the same-identification road segment when the number of trajectories is not less than the first preset trajectory threshold, to obtain a target trajectory. The fifth fusion sub-module is configured to return to the steps of obtaining the driving trajectory of the target vehicle, the road information, and the navigation information when the number of trajectories is less than the first preset trajectory threshold.

[0158] In some optional embodiments, the fourth fusion sub-module includes a first fusion unit, a second fusion unit, and a third fusion unit. The first fusion unit is configured to map the driving trajectories in each same-identification road segment to a same coordinate system. The second fusion unit is configured to perform fitting degree calculation on all the driving trajectories in each same-identification road segment, and perform trajectory fusion based on the fitting degree calculation result to obtain a fused trajectory. The third fusion unit is configured to integrate the fused trajectories corresponding to all the same-identification road segments to obtain a target trajectory.

[0159] In some optional embodiments, the second fusion unit includes a first fusion sub-unit, a second fusion sub-unit, a third fusion sub-unit, a fourth fusion sub-unit, a fifth fusion sub-unit, and a sixth fusion sub-unit. The first fusion sub-unit is configured to calculate a lateral distance between two adjacent driving trajectories in each same-identification road segment. The second fusion sub-unit is configured to determine a fitting degree calculation result of the two adjacent driving trajectories based on the lateral distance. The third fusion sub-unit is configured to filter all the driving trajectories with a fitting degree calculation result greater than a preset fitting degree threshold, and construct a trajectory dataset. The fourth fusion sub-unit is configured to determine whether a number of trajectories in the trajectory dataset is less than a second preset trajectory threshold. The fifth fusion sub-unit is configured to perform trajectory fusion on the trajectory dataset when the number of trajectories is not less than the second preset trajectory threshold, to obtain a fused trajectory. The sixth fusion sub-unit is configured to return to the steps of obtaining the driving trajectory of the target vehicle, the road information, and the navigation information when the number of trajectories is less than the second preset trajectory threshold.

[0160] In some optional embodiments, the extraction module 903 comprises a first extraction submodule, a second extraction submodule and a third extraction submodule; the first extraction submodule is configured to filter road information corresponding to the target trajectory; the second extraction submodule is configured to extract target scenes from the road information based on preset scene elements, wherein the preset scene elements comprise lane lines, buildings and road structures; and the third extraction submodule is configured to integrate all the target scenes to obtain a scene dataset.

[0161] In some optional embodiments, the production module 904 comprises a first production submodule and a second production submodule; the first production submodule is configured to associate all the target scenes in the scene dataset with the target trajectory respectively to obtain an associated trajectory of the target vehicle; and the second production submodule is configured to generate a map of the target vehicle based on the associated trajectory.

[0162] In some optional embodiments, the system further comprises a verification submodule configured to generate a standard definition map based on the navigation information; determine whether the target scenes in the scene dataset match corresponding positions in the standard definition map respectively; retain the corresponding target scenes when the target scenes match the corresponding positions in the standard definition map; and delete the corresponding target scenes and update the scene dataset when the target scenes do not match the corresponding positions in the standard definition map.

[0163] Further function descriptions of the above modules are the same as those of the corresponding embodiments, and will not be repeated here.

[0164] The lightweight map production system in the present embodiment is presented in the form of functional units, wherein the units refer to ASIC (Application Specific Integrated Circuit) circuits, processors and memories executing one or more software or fixed programs, and / or other devices capable of providing the above functions.

[0165] The lightweight map production system of the present embodiment can greatly reduce the data amount required for producing a map, improve the accuracy and efficiency of map production, reduce the update cycle, and meet the intelligent navigation requirements of a vehicle by fusing trajectories and extracting scenes based on the trajectories and road conditions during the driving process of the vehicle.

[0166] The present embodiment further provides a vehicle comprising a controller. The controller in the present embodiment is a vehicle controller for performing power-on / power-off, hibernation and wake-up operations on the sub-controllers and network nodes hung thereunder, and each power supply interface of the controller can collect real-time current output. Other controllers with the above functions are also applicable.

[0167] Figure 10 is a structural schematic diagram of the controller provided by the optional embodiment of the present application, as shown in the figure, the controller comprises one or more processors 10, a memory 20, and an interface for connecting various components, including a high-speed interface and a low-speed interface. Various components are communicatively connected to each other by different buses, and can be installed on a common mainboard or otherwise installed as needed. The processor can process instructions executed within the controller, including instructions stored in the memory or on the memory to display graphical information of a GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, multiple processors and / or multiple buses can be used with multiple memories and multiple memories, if needed. Similarly, multiple controllers can be connected, each providing part of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 10 The processor 10 is taken as an example in the figure. Figure 10

[0168] The processor 10 can be a central processor, a network processor, or a combination thereof. The processor 10 can further include a hardware chip. The hardware chip can be an application specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device can be a complex programmable logic device, a field programmable logic gate array, a generic array logic, or any combination thereof.

[0169] The memory 20 stores instructions executable by the at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0170] The memory 20 can include a program storage area and a data storage area. The program storage area can store an operating system and application programs required by at least one function; the data storage area can store data created according to the use of the controller, etc. In addition, the memory 20 can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some optional embodiments, the memory 20 can optionally include a memory remotely disposed relative to the processor 10, which can be connected to the controller through a network. Examples of the network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0171] The memory 20 can include a volatile memory, such as a random access memory; the memory can also include a non-volatile memory, such as a flash memory, a hard disk, or a solid state disk; the memory 20 can also include a combination of the above types of memories.

[0172] ​The controller also includes a communication interface 30 for the master chip to communicate with other devices or communication networks.

[0173] The embodiments of the present application also provide a computer readable storage medium, the method according to the embodiments of the present application can be implemented in hardware, firmware, or be implemented as computer code recorded in a storage medium, or be implemented through network downloading and originally stored in a remote storage medium or a non-transitory machine readable storage medium and to be stored in a local storage medium, so that the method described herein can be processed by such software on a storage medium using a general purpose computer, a special purpose processor or programmable or special purpose hardware. Wherein, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk or a solid state disk, etc. Further, the storage medium can also include a combination of the above-mentioned types of memories. It can be understood that the computer, the processor, the microprocessor master chip or the programmable hardware includes a storage component that can store or receive software or computer code, when the software or computer code is accessed and executed by the computer, the processor or the hardware, the method shown in the above embodiments is implemented.

[0174] Although the embodiments of the present application are described in conjunction with the accompanying drawings, various modifications and changes can be made by those skilled in the art without departing from the spirit and scope of the present application, and such modifications and changes fall within the scope defined by the appended claims.

Claims

1. A light-weight map production method characterized by comprising: The method comprises: obtaining a driving track of a target vehicle, road information and navigation information; fusing the driving track based on the navigation information and the road information to obtain a target track; extracting a target scene from the road information to construct a scene dataset; making a map of the target vehicle based on the target track and the scene dataset.

2. The light-weight map generation method according to claim 1, characterized by, The fusing of the driving track based on the navigation information and the road information to obtain a target track comprises: segmenting the road information based on the road structure of the navigation information to obtain a plurality of road segments and identifying each road segment; selecting, for each road segment, driving tracks with the same road segment identification to obtain a plurality of same-identification road segments; judging whether the number of tracks in each same-identification road segment is less than a first preset track threshold; when the number of tracks is not less than the first preset track threshold, fusing the driving tracks in the corresponding same-identification road segment to obtain a target track; when the number of tracks is less than the first preset track threshold, returning to the step of obtaining the driving track of the target vehicle, the road information and the navigation information.

3. The light-weight map generation method according to claim 2, characterized by, The fusing of the driving tracks in the corresponding same-identification road segment to obtain a target track comprises: mapping the driving tracks in each same-identification road segment to the same coordinate system; calculating the fitting degree of all the driving tracks in each same-identification road segment and fusing the tracks based on the fitting degree calculation result to obtain a fused track; integrating the fused tracks corresponding to all the same-identification road segments to obtain a target track.

4. The light-weight map generation method according to claim 3, characterized by, The calculation of the fitting degree of all the driving tracks in each same-identification road segment and the fusing of the tracks based on the fitting degree calculation result to obtain a fused track comprises: calculating the lateral distance between two adjacent driving tracks for each same-identification road segment; determining the fitting degree calculation result of the two adjacent driving tracks based on the lateral distance; selecting all the driving tracks with the fitting degree calculation result greater than a preset fitting degree threshold to construct a track dataset; judging whether the number of tracks in the track dataset is less than a second preset track threshold; when the number of tracks is not less than the second preset track threshold, fusing the tracks in the track dataset to obtain a fused track; when the number of tracks is less than the second preset track threshold, returning to the step of obtaining the driving track of the target vehicle, the road information and the navigation information.

5. The lightweight map production method according to claim 1, characterized by, The extraction of a target scene from the road information to construct a scene dataset comprises: selecting road information corresponding to a target track; extracting a target scene from the road information based on preset scene elements, the preset scene elements including lane lines, buildings and road structures; integrating all the target scenes to obtain a scene dataset.

6. The lightweight map generation method according to any one of claims 1 to 5, characterized by, The making of a map of the target vehicle based on the target track and the scene dataset comprises: associating all the target scenes in the scene dataset with the target track to obtain an associated track of the target vehicle; generating a map of the target vehicle based on the associated track.

7. The light-weight map generation method according to claim 6, characterized by, Before making the map of the target vehicle based on the target trajectory and the scene data set, the method further comprises: generating a standard definition map based on navigation information; respectively judging whether a target scene in the scene data set matches a corresponding position in the standard definition map; when the target scene matches the corresponding position in the standard definition map, retaining the corresponding target scene; when the target scene does not match the corresponding position in the standard definition map, deleting the corresponding target scene and updating the scene data set.

8. A lightweight map production system characterized by comprising: The system comprises: an acquisition module configured to acquire a driving trajectory of a target vehicle, road information and navigation information; a fusion module configured to perform fusion processing on the driving trajectory based on the navigation information and the road information to obtain a target trajectory; an extraction module configured to extract a target scene from the road information to construct a scene data set; a making module configured to make a map of the target vehicle based on the target trajectory and the scene data set.

9. A vehicle characterized by comprising: The vehicle comprises a controller, the controller comprising a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the lightweight map making method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions, and the computer instructions are used to make the computer execute the lightweight map making method of any one of claims 1 to 7.

Images