Track deviation correction method and device, computer device and storage medium

Abstract

The application relates to a trajectory deviation correction method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: determining a target deviation correction reference point according to a preset first selection strategy; determining a first subset of collected trajectory points according to a preset deviation correction selection condition; determining a first subset of initial deviation correction trajectory points based on trajectory data of the target deviation correction reference point, a subset of arrival times corresponding to the first subset of collected trajectory points, and a pre-trained position deviation correction model; determining the deviation correction accuracy of the first subset of initial deviation correction trajectory points according to the first subset of initial deviation correction trajectory points and an actual matching lane data subset; in the case that the deviation correction accuracy does not meet a preset deviation correction accuracy condition, the target deviation correction reference point is determined again, and the step of determining the target deviation correction reference point according to the preset deviation correction selection condition is executed again until the deviation correction accuracy meets the preset accuracy condition, and a first subset of deviation correction trajectory points after deviation correction is obtained. The scheme improves the deviation correction accuracy of trajectory deviation correction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of trajectory data processing, in particular to a trajectory correction method and device, computer equipment and a storage medium. BACKGROUND

[0002] With the development of trajectory data processing technology, trajectory correction technology appears. This technology can correct low-precision trajectory data collected by a low-precision global positioning system (GPS) device, and then obtain corrected trajectory data. The corrected trajectory data can provide a data source for road condition analysis, map drawing, etc.

[0003] In the traditional technology, the target trajectory data of the trajectory point at the current time and the arrival time of the trajectory point at the next time are input into a pre-trained position prediction model to obtain the position data of the trajectory point at the next time. Then, the trajectory data of the trajectory point at the next time collected by the global positioning system device is obtained, and the position data in the initial trajectory data is corrected based on the position data of the trajectory point to obtain the trajectory data of the trajectory point at the next time (i.e. the corrected trajectory data). Assuming that the target vehicle generates x trajectory points in a certain driving, after (x-1) times of trajectory correction, (x-1) corrected trajectory data are obtained.

[0004] However, the correction reference points in the current trajectory correction technology are all the target trajectory data of the last trajectory point. However, if the correction quality of the target trajectory data of the last trajectory point is poor, the correction quality of the target trajectory data of the subsequent trajectory points will become poorer and poorer, thereby resulting in low correction accuracy of trajectory correction. SUMMARY

[0005] Therefore, it is necessary to provide a trajectory correction method, device, computer equipment, computer readable storage medium and computer program product capable of improving correction accuracy in view of the above technical problems.

[0006] In a first aspect, the present application provides a trajectory correction method. The method comprises:

[0007] determining a target correction reference point in a correction reference point set according to a preset first selection strategy;

[0008] determining a first subset of collected trajectory points in the collected trajectory point set according to a preset correction selection condition;

[0009] determining a first subset of initial correction trajectory points corresponding to the first subset of collected trajectory points based on the trajectory data of the target correction reference point, a subset of arrival times corresponding to the first subset of collected trajectory points, and a pre-trained position correction model.

[0010] determine the rectification accuracy of the first subset of initial rectification trajectory points according to the first subset of initial rectification trajectory points and the subset of actual matching lane data corresponding to the first subset of initial rectification trajectory points;

[0011] In a case where the rectification accuracy does not satisfy the preset rectification accuracy condition, the target rectification reference point is re-determined in the set of rectification reference points based on the preset first selection strategy, and the step of determining the first subset of initial rectification trajectory points according to the preset rectification selection condition is executed again until the rectification accuracy satisfies the preset accuracy condition, and the first subset of initial rectification trajectory points is taken as the first subset of rectification trajectory points after rectification.

[0012] In one of the embodiments, before the target rectification reference point is determined in the set of rectification reference points according to the preset first selection strategy, the method further comprises:

[0013] determine the target historical trajectory points in the set of historical trajectory points according to the preset second selection strategy to obtain a set of target historical trajectory points;

[0014] perform lane matching based on the trajectory data of the target historical trajectory points and the lane data of the lane center line to determine the predicted matching lane of the target historical trajectory points;

[0015] determine the collection accuracy of the trajectory data of the target historical trajectory points based on the predicted matching lane of the target historical trajectory points and the actual matching lane of the target historical trajectory points, and determine the collection accuracy of the set of target historical trajectory points according to the collection accuracy of each target historical trajectory point;

[0016] In a case where the collection accuracy of the set of target historical trajectory points does not satisfy the preset collection accuracy condition, the target historical trajectory points are re-determined in the set of historical trajectory points based on the preset second selection strategy, and the step of obtaining the set of target historical trajectory points is executed again until the collection accuracy of the set of target historical trajectory points satisfies the preset collection accuracy condition, and the set of rectification reference points is determined according to the set of target historical trajectory points.

[0017] In one of the embodiments, the determining the first subset of collection trajectory points in the set of collection trajectory points according to the preset rectification selection condition comprises:

[0018] In a case where the arrival time of a collection trajectory point is the same as the shooting time of a vehicle-mounted image in the set of collection trajectory points, the collection trajectory point is taken as a diversity trajectory point, and the arrival time of the diversity trajectory point is taken as a diversity time;

[0019] In the collection trajectory point set, a collection trajectory point with an arrival time earlier than or equal to a first distribution time is determined to obtain a first collection trajectory point subset; the first distribution time is the earliest distribution time.

[0020] In one of the embodiments, the determining of the correction accuracy of the initial correction trajectory point first subset comprises:

[0021] According to a preset third selection strategy, a target correction trajectory point is determined in the initial correction trajectory point first subset, and correction trajectory data of the target correction trajectory point is obtained;

[0022] Lane matching is performed based on the correction trajectory data of the target correction trajectory point and lane data of the lane center line to determine a predicted matching lane of the target correction trajectory point.

[0023] Based on the predicted matching lane of the target correction trajectory point and an actual matching lane of the target correction trajectory point, the correction accuracy of the target correction trajectory point is determined, and the correction accuracy of the initial correction trajectory point first subset is determined according to the correction accuracy of each target correction trajectory point.

[0024] In one of the embodiments, the determining of the correction reference point set according to the target historical trajectory point set comprises:

[0025] For each target historical trajectory point in the target historical trajectory point set, if the collection accuracy of the target historical trajectory point meets the preset collection accuracy condition, the target historical trajectory point is taken as the correction reference point.

[0026] In one of the embodiments, after the initial correction trajectory point first subset is taken as the corrected correction trajectory point first subset, the method further comprises:

[0027] The corrected correction trajectory point first subset is taken as a new correction reference point set, and the step of determining a target correction reference point in the correction reference point set according to the preset first selection strategy is performed.

[0028] In one of the embodiments, the determining of the target correction reference point in the correction reference point set according to the preset first selection strategy comprises:

[0029] From the correction reference point set, a correction reference point corresponding to the latest arrival time is selected as the target correction reference point.

[0030] In a second aspect, the application further provides a trajectory correction device. The device comprises:

[0031] The first determining module is configured to determine a target rectification reference point from the set of rectification reference points according to a preset first selection strategy.

[0032] The second determining module is configured to determine a first subset of collected trajectory points from the set of collected trajectory points according to a preset rectification selection condition.

[0033] The third determining module is configured to determine a first subset of initial rectified trajectory points corresponding to the first subset of collected trajectory points based on trajectory data of the target rectification reference point, a subset of arrival times corresponding to the first subset of collected trajectory points, and a pre-trained position rectification model.

[0034] The fourth determining module is configured to determine rectification accuracy of the first subset of initial rectified trajectory points according to the first subset of initial rectified trajectory points and a subset of actual matching lane data corresponding to the first subset of initial rectified trajectory points.

[0035] The fifth determining module is configured to, in a case where the rectification accuracy does not satisfy a preset rectification accuracy condition, re-determine the target rectification reference point from the set of rectification reference points based on the preset first selection strategy, and return to perform the step of determining the target rectification reference point according to the preset rectification selection condition until the rectification accuracy satisfies the preset accuracy condition, and take the first subset of initial rectified trajectory points as a first subset of rectified trajectory points.

[0036] In one of the embodiments, the trajectory rectification apparatus further comprises:

[0037] The sixth determining module is configured to determine a target historical trajectory point from the set of historical trajectory points according to a preset second selection strategy, to obtain a set of target historical trajectory points.

[0038] The seventh determining module is configured to perform lane matching based on trajectory data of the target historical trajectory point and lane data of a lane center line, to determine a predicted matching lane of the target historical trajectory point.

[0039] The eighth determining module is configured to determine collection accuracy of trajectory data of the target historical trajectory point based on the predicted matching lane of the target historical trajectory point and an actual matching lane of the target historical trajectory point, and determine collection accuracy of the set of target historical trajectory points according to collection accuracy of each target historical trajectory point.

[0040] the ninth determining module is configured to, in a case where the collection accuracy of the target historical trajectory point set does not satisfy a preset collection accuracy condition, determine the target historical trajectory point again in the historical trajectory point set based on a preset second selection strategy, and return to perform the step of obtaining the target historical trajectory point set until the collection accuracy of the target historical trajectory point set satisfies the preset collection accuracy condition, and determine the set of correction reference points based on the target historical trajectory point set.

[0041] In one of the embodiments, the second determining module is specifically configured to:

[0042] In a case where the arrival time of a collection trajectory point is the same as the shooting time of the vehicle-mounted image, the collection trajectory point is taken as a diversity trajectory point, and the arrival time of the diversity trajectory point is taken as a diversity time.

[0043] In the collection trajectory point set, a collection trajectory point with an arrival time earlier than or equal to a first diversity time is determined to obtain a first collection trajectory point subset; the first diversity time is the earliest diversity time.

[0044] In one of the embodiments, the fourth determining module is specifically configured to:

[0045] According to a preset third selection strategy, a target correction trajectory point is determined in the initial correction trajectory point first subset, and correction trajectory data of the target correction trajectory point is obtained;

[0046] Based on the correction trajectory data of the target correction trajectory point and lane data of the lane center line, lane matching is performed to determine a predicted matching lane of the target correction trajectory point.

[0047] Based on the predicted matching lane of the target correction trajectory point and the actual matching lane of the target correction trajectory point, correction accuracy of the target correction trajectory point is determined, and correction accuracy of the initial correction trajectory point first subset is determined according to the correction accuracy of each target correction trajectory point.

[0048] In one of the embodiments, the ninth determining module is specifically configured to:

[0049] In a case where the collection accuracy of each target historical trajectory point in the target historical trajectory point set satisfies the preset collection accuracy condition, the target historical trajectory point is taken as the correction reference point.

[0050] In one of the embodiments, the trajectory correction device further includes:

[0051] The circulation module is configured to take the deviated trajectory point first subset after correction as a new set of deviated reference points, and return to execute the step of determining a target deviated reference point from the set of deviated reference points according to the preset first selection strategy.

[0052] In one of the embodiments, the first determining module is specifically configured to:

[0053] The target deviated reference point is selected from the set of deviated reference points as a deviated reference point corresponding to the latest arrival time.

[0054] In a third aspect, the present application further provides a computer device. The computer device comprises a memory and a processor. The memory stores a computer program. The processor executes the computer program to implement the steps of the first aspect.

[0055] In a fourth aspect, the present application further provides a computer readable storage medium. The computer readable storage medium stores a computer program. The computer program is executed by a processor to implement the steps of the first aspect.

[0056] In a fifth aspect, the present application further provides a computer program product. The computer program product comprises a computer program. The computer program is executed by a processor to implement the steps of the first aspect.

[0057] The trajectory deviation correction method, device, computer device, storage medium and computer program product can determine a target deviated reference point from the set of deviated reference points according to the preset first selection strategy, determine a deviated trajectory point first subset from the set of collected trajectory points according to the preset deviation selection condition, determine an initial deviated trajectory point first subset corresponding to the deviated trajectory point first subset based on the trajectory data of the target deviated reference point, an arrival time subset corresponding to the deviated trajectory point first subset and a pre-trained position deviation correction model, determine the deviation correction accuracy of the initial deviated trajectory point first subset according to the initial deviated trajectory point first subset and an actual matching lane data subset corresponding to the initial deviated trajectory point first subset, and in the case that the deviation correction accuracy does not meet the preset deviation correction accuracy condition, re-determine the target deviated reference point from the set of deviated reference points based on the preset first selection strategy, and return to execute the step of determining the target deviated reference point from the set of deviated reference points according to the preset deviation selection condition until the deviation correction accuracy meets the preset accuracy condition, and take the initial deviated trajectory point first subset as a deviated trajectory point first subset after correction. It can be understood that the initial deviated trajectory point first subset is determined based on the set of deviated reference points, and in the case that the deviation correction accuracy of the initial deviated trajectory point first subset meets the preset deviation correction accuracy condition, the initial deviated trajectory point first subset is taken as the deviated trajectory point first subset after correction. Therefore, the deviation correction quality of the initial deviated trajectory point first subset can be ensured, and the deviation correction accuracy of the trajectory deviation correction is improved. Attached Figure Description

[0058] Figure 1 This is a flowchart illustrating a trajectory correction method in one embodiment;

[0059] Figure 2 This is a flowchart illustrating a method for obtaining a set of correction reference points in one embodiment;

[0060] Figure 3 This is a schematic diagram of a lane-level map in one embodiment;

[0061] Figure 4 This is a flowchart illustrating a method for acquiring the first subset of trajectory points in one embodiment;

[0062] Figure 5 This is a structural block diagram of a trajectory correction device in one embodiment;

[0063] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0065] In one embodiment, such as Figure 1 As shown, a trajectory correction method is provided. This embodiment illustrates the application of this method to a terminal. It is understood that this method can also be applied to a server, and further to a system including both a terminal and a server, and implemented through interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0066] Step 102: Determine the target correction reference point from the set of correction reference points according to the preset first selection strategy.

[0067] The set of rectification reference points includes at least one rectification reference point. The rectification reference point is a historical trajectory point of the terminal, and is used to rectify a first collected trajectory point of the terminal in a driving trajectory. The collected trajectory point refers to a trajectory point corresponding to trajectory data collected by the terminal in a rectification environment. That is, the collected trajectory point is a trajectory point actually generated when the terminal moves in a rectification environment. The trajectory data corresponding to the trajectory point includes a longitude of the trajectory point, a latitude of the trajectory point, an altitude of the terminal, a speed of the terminal, a collection time of a global positioning system (GPS) device (that is, a time when the terminal reaches the trajectory point, which is referred to as a reaching time for the sake of convenience), and a direction of the terminal. The first collected trajectory point refers to a collected trajectory point corresponding to an earliest reaching time in a driving.

[0068] In the embodiment of the present application, the terminal obtains the set of rectification reference points, and selects one of the rectification reference points as a target rectification reference point according to a preset first selection strategy. In one embodiment, the terminal randomly selects one of the rectification reference points as the target rectification reference point. In another embodiment, the terminal obtains trajectory data of the first collected point, and calculates a selection value for each rectification reference point in the set of rectification reference points according to the trajectory data of the first collected point, trajectory data of the rectification reference point, and a preset first selection formula. Then, the terminal filters the selection values according to the preset first selection strategy to obtain a target selection value, and takes the rectification reference point corresponding to the target selection value as the target rectification reference point. The preset first selection formula is a weighted summation formula constructed based on the trajectory data. Specifically, the preset first selection formula can be y n =k1·t n +k2·d n +k3·h n +k4·v n +k5·a n , where y n is the selection value of the nth rectification reference point, k1, k2, k3, k4, and k5 are preset weights, k1+k2+k3+k4+k5=1, t n is a reaching time difference between the nth rectification reference point and the first collected point, d n is a relative distance between the nth rectification reference point and the first collected point, h n is the altitude of the nth rectification reference point, v n is the speed of the nth rectification reference point, a n is an azimuth angle between the nth rectification reference point and the first collected point, and n is an integer greater than 0. In one embodiment, k3=k4=k5=0, that is, the preset first selection formula is y n=k1·t n +k2·d n The terminal acquires the trajectory data of the first collection point, and determines the longitude and latitude of the first collection point, and the arrival time. For each rectification reference point in the rectification reference point set, the terminal calculates the relative distance between the first collection point and the rectification reference point according to the longitude and latitude of the first collection point and the longitude and latitude of the rectification reference point. The terminal calculates the arrival time difference between the first collection point and the rectification reference point according to the arrival time of the first collection point and the arrival time of the rectification reference point. The terminal calculates the selected value y n =k1·t n +k2·d n according to the preset first selection formula y n = k1·t + k2·d, and takes the rectification reference point corresponding to the smallest selected value as the target rectification reference point.

[0069] Step 104, determining a first subset of collection trajectory points in the collection trajectory point set according to a preset rectification selection condition.

[0070] The collection trajectory point set includes at least one collection trajectory point. The collection trajectory point refers to the trajectory point corresponding to the trajectory data collected by the terminal in the rectification environment. That is, the collection trajectory point is the trajectory point actually generated when the terminal moves in the rectification environment. The first subset of collection trajectory points is a subset of the collection trajectory point set, and includes the arrival time of the sampling trajectory point earlier than or equal to the first subset time. The subset time is the arrival time of the subset trajectory point. The first subset time is the earliest subset time. The subset trajectory point is the collection trajectory point used to determine the subset of collection trajectory points.

[0071] In the embodiments of the present application, for each collection trajectory point in the collection trajectory point set, the terminal compares the arrival time of the collection trajectory point with the first diversity time, and takes the trajectory point with an arrival time earlier than or equal to the first diversity time as an element in the first subset of collection trajectory points, to obtain the first subset of collection trajectory points. For example, the collection trajectory point set is {x1, x2, x3, x4, x5, x6}, where the collection trajectory point set is an ordered set, and each collection trajectory point in the collection trajectory point set is sorted according to the arrival time, which can be from left to right, and the arrival time of the collection trajectory point is from early to late. Assuming that x3 and x5 are diversity trajectory points, because the arrival time of the diversity trajectory point x3 is earlier than the arrival time of the diversity trajectory point x5, the arrival time of the diversity trajectory point x3 is the first diversity time, and the arrival time of the diversity trajectory point x5 is the second diversity time. For the collection trajectory point set {x1, x2, x3, x4, x5, x6}, the terminal compares the arrival time of the collection trajectory point with the arrival time of the diversity trajectory point x3 (i.e., the first diversity time), and obtains that the arrival time of the collection trajectory point x1 is earlier than the first diversity time, the arrival time of the collection trajectory point x2 is earlier than the first diversity time, and the arrival time of the collection trajectory point x3 is equal to the first diversity time. Therefore, the terminal takes the collection trajectory point x1, the collection trajectory point x2, and the collection trajectory point x3 as elements in the first subset of collection trajectory points, to obtain the first subset of sampling trajectory points {x1, x2, x3}.

[0072] In step 106, based on the trajectory data of the target deviation correction reference point, the arrival time subset corresponding to the first subset of collection trajectory points, and the pre-trained position deviation correction model, the first subset of initial deviation correction trajectory points corresponding to the first subset of collection trajectory points is determined.

[0073] The trajectory data of the target deviation correction reference point includes the longitude of the target deviation correction reference point, the latitude of the target deviation correction reference point, the elevation of the terminal, the speed of the terminal, the collection time of the global positioning system (GPS) device (i.e., the time when the terminal arrives at the target deviation correction reference point, which is referred to as the arrival time of the target deviation correction reference point for convenience of distinction), and the direction of the terminal. The arrival time subset corresponding to the first subset of collection trajectory points includes the arrival time of at least one collection trajectory point, where the collection trajectory point is an element in the first subset of collection trajectory points. For example, the first subset of sampling trajectory points is {x1, x2, x3}, and assuming that t1 is the arrival time of the collection trajectory point x1, t2 is the arrival time of the collection trajectory point x2, and t3 is the arrival time of the collection trajectory point x3, the arrival time subset corresponding to the first subset of collection trajectory points is {t1, t2, t3}.

[0074] In the embodiment of the present application, the terminal performs lane matching processing according to the trajectory data of the target deviation correction reference point and the lane data of the lane center line, and obtains the predicted matching lane of the target deviation correction reference point. The lane data of the lane center line includes the position data of the lane center line and the lane identifier of the lane where the lane center line is located. The terminal performs deviation correction processing according to the trajectory data of the target deviation correction reference point and the predicted matching lane of the target deviation correction reference point, and obtains the reference trajectory data of the target deviation correction reference point. Specifically, the terminal corrects the longitude and latitude in the trajectory data of the target deviation correction reference point to the predicted matching lane according to the predicted matching lane of the target deviation correction reference point, and obtains the reference trajectory data. The terminal inputs the reference trajectory data of the target deviation correction reference point and the arrival time of the first collected trajectory point in the first subset of collected trajectory points into the pre-trained position correction model, and outputs the predicted longitude and latitude of the first collected trajectory point. The position correction model can be a Hidden Markov Model (HMM), an Interacting Multiple Model (IMM), or a Kalman filtering. The terminal updates the predicted longitude and latitude of the first collected trajectory point to the longitude and latitude in the trajectory data of the first collected trajectory point, obtains the initial deviation correction trajectory data, and takes the first collected trajectory point as the first initial deviation correction trajectory point.

[0075] In one embodiment, for the target deviation correction reference point, the terminal performs data supplement processing on the reference trajectory data according to the pre-set data type of the target trajectory data, and obtains the target reference trajectory data of the target trajectory data. The data type of the target reference trajectory data of the target deviation correction reference point includes the data type of the reference trajectory data and the supplement data type. In one embodiment, the supplement data type of the target deviation correction reference point includes the lane identifier of the predicted matching lane of the target deviation correction reference point and the relative distance between the target deviation correction reference point and the predicted matching lane. The terminal inputs the target reference trajectory data of the target deviation correction reference point and the arrival time of the first collected trajectory point in the first subset of collected trajectory points into the pre-trained position correction model, and outputs the predicted longitude and latitude of the first collected trajectory point. The terminal updates the predicted longitude and latitude of the first collected trajectory point to the longitude and latitude in the trajectory data of the first collected trajectory point, obtains the initial deviation correction trajectory data, and takes the first collected trajectory point as the first initial deviation correction trajectory point.

[0076] The method of obtaining the (N+1)th initial rectification trajectory point based on the initial rectification trajectory data of the Nth initial rectification trajectory point, the arrival time of the (N+1)th sampling trajectory point in the first subset of sampling trajectory points, and the pre-trained position rectification model is similar to the method of obtaining the first initial rectification trajectory point based on the trajectory data of the target rectification reference point, the arrival time of the first sampling trajectory point in the first subset of sampling trajectory points, and the pre-trained position rectification model, and thus will not be described again. If the first subset of sampling trajectory points has m sampling trajectory points, after m times of the above rectification processing, m initial rectification trajectory points are obtained, and the first subset of initial rectification trajectory points is formed based on the m initial rectification trajectory points. For example, assuming that the first subset of sampling trajectory points is {x1, x2, x3}, the first subset of initial rectification trajectory points corresponding to the first subset of sampling trajectory points is {x'1, x'2, x'3}. The initial rectification trajectory point x'1 is obtained by rectifying the first sampling trajectory point x1, the initial rectification trajectory point x'2 is obtained by rectifying the second sampling trajectory point x2, and the initial rectification trajectory point x'3 is obtained by rectifying the third sampling trajectory point x3.

[0077] In step 108, the rectification accuracy of the first subset of initial rectification trajectory points is determined according to the first subset of initial rectification trajectory points and the subset of actual matching lane data corresponding to the first subset of initial rectification trajectory points.

[0078] In the embodiment of the present application, the terminal selects a first preset number of initial rectification trajectory points from the first subset of initial rectification trajectory points as target rectification trajectory points according to a preset third selection strategy. For each target rectification trajectory point, the terminal obtains the rectification accuracy of the target rectification trajectory point according to the predicted matching lane of the target rectification trajectory point and the actual matching lane of the target rectification trajectory point. The terminal calculates the rectification accuracy of the first subset of initial rectification trajectory points based on the rectification accuracy of the first preset number of target rectification trajectory points. In one embodiment, the terminal calculates the average of the rectification accuracy of the first preset number of target rectification trajectory points, and takes the average as the rectification accuracy of the first subset of initial rectification trajectory points. In one embodiment, the first preset number is 2, and the target rectification trajectory points are the initial rectification trajectory point with the earliest arrival time and the initial rectification trajectory point with the latest arrival time in the first subset of initial rectification trajectory points.

[0079] In step 110, in the case where the rectification accuracy does not satisfy the preset rectification accuracy condition, the target rectification reference point is re-determined in the rectification reference point set based on the preset first selection strategy, and the step of determining the target rectification reference point according to the preset rectification selection condition is executed again until the rectification accuracy satisfies the preset accuracy condition, and the first subset of initial rectification trajectory points is taken as the first subset of rectified rectification trajectory points.

[0080] In the case where the correction accuracy of the initial correction trajectory point first subset does not meet the preset correction accuracy condition, the terminal selects, according to a preset first selection strategy, a correction reference point that has not been selected as a target correction reference point from the correction reference point set as the target correction reference point, and returns to perform step 104 until the correction accuracy of the initial correction trajectory point first subset meets the preset accuracy condition, and the terminal takes the initial correction trajectory point first subset as the corrected correction trajectory point first subset. It can be understood that the method of selecting, according to the preset first selection strategy, a correction reference point that has not been selected as a target correction reference point from the correction reference point set as the target correction reference point is similar to the method of step 102, and will not be described again. In an embodiment, the preset correction accuracy condition is that the correction accuracy of the initial correction trajectory point first subset is 1.

[0081] In the trajectory correction method, the initial correction trajectory point first subset is determined based on the correction reference point set, and the initial correction trajectory point first subset is taken as the corrected correction trajectory point first subset only in the case where the correction accuracy of the initial correction trajectory point first subset meets the preset correction accuracy condition. Therefore, the correction quality of the initial correction trajectory point first subset can be guaranteed, and the correction accuracy of the trajectory correction is improved.

[0082] In an embodiment, as shown in FIG. 2, before determining the target correction reference point from the correction reference point set according to the preset first selection strategy, the method further includes: Figure 2

[0083] Step 202: determining a target historical trajectory point from the historical trajectory point set according to a preset second selection strategy, to obtain a target historical trajectory point set.

[0084] The historical trajectory point set includes at least one historical trajectory point. The historical trajectory point is a trajectory point corresponding to trajectory data collected by the terminal at a historical time. The target historical trajectory point set includes at least one target historical trajectory point.

[0085] ​In the embodiment of the present application, the terminal obtains a historical trajectory point set, and selects a historical trajectory point as a target historical trajectory point according to a preset second selection strategy in the historical trajectory point set, to obtain a target historical trajectory point set. Specifically, for each historical trajectory point in the historical trajectory point set, the terminal obtains an arrival time corresponding to the historical trajectory point, and searches for whether there is a shooting time identical to the arrival time corresponding to the historical trajectory point. In the case that the shooting time of the vehicle-mounted image is identical to the arrival time corresponding to the historical trajectory point, the terminal takes the historical trajectory point as an initial historical trajectory point. The terminal takes a second preset number of initial historical trajectory points as target historical trajectory points to construct a set, to obtain the target historical trajectory point set. In an embodiment, the terminal sorts the initial historical trajectory points according to the arrival times of the initial historical trajectory points, to obtain an initial historical trajectory point sequence, and randomly selects a second preset number of initial historical trajectory points from the initial historical trajectory point sequence as target historical trajectory points to construct a set, to obtain the target historical trajectory point set.

[0086] In step 204, lane matching is performed based on the trajectory data of the target historical trajectory point and the lane data of the lane center line, to determine a predicted matching lane of the target historical trajectory point.

[0087] The lane data of the lane center line includes position data of the lane center line and a lane identifier of a lane where the lane center line is located.

[0088] In the embodiment of the present application, the terminal obtains a GPS map pre-stored in the terminal, and performs positioning on the GPS map according to the trajectory data of the target historical trajectory point, to obtain lane data of two lane center lines adjacent to the target historical trajectory point. The terminal performs lane matching according to the trajectory data of the target historical trajectory point and the position data of the two lane center lines, to obtain a predicted matching lane of the target historical trajectory point. Specifically, the terminal calculates a relative distance between the target historical trajectory point and the lane center line (for the convenience of distinction, it is called point-lane distance) according to the latitude and longitude in the trajectory data of the target historical trajectory point and the position data of the lane center line, and takes a lane where the lane center line corresponding to the smallest point-lane distance is located as the predicted matching lane of the target historical trajectory point. For example, Figure 3As shown, assuming that L1 and L2 represent lane center lines adjacent to the target historical trajectory point on the GPS map, the black solid point A represents the target historical trajectory point A, d1 represents the relative distance between the target historical trajectory point A and the lane center line L1, and d2 represents the relative distance between the target historical trajectory point A and the lane center line L2. In the case where d1 is greater than d2, the terminal takes the lane in which the lane center line L1 is located as the predicted matching lane of the target historical trajectory point A. In the case where the two lane distances are equal, the terminal obtains the lane level of the lane in which the lane center line is located according to the lane identification of the lane in which the lane center line is located and the preset lane level, and takes the lane with a higher lane level as the predicted matching lane of the target historical trajectory point. Optionally, the preset lane level can be an administrative level type of road level (national highway, provincial highway, county highway, and township highway as well as special-purpose highway, from left to right, and the level from high to low), or a functional type of road level (expressway, first-class highway, second-class highway, third-class highway, and fourth-class highway, from left to right, and the level from high to low).

[0089] In step 206, the collection accuracy of the target historical trajectory point is determined based on the predicted matching lane of the target historical trajectory point and the actual matching lane of the target historical trajectory point, and the collection accuracy of the target historical trajectory point set is determined according to the collection accuracy of each target historical trajectory point.

[0090] In the embodiments of the present application, for each target historical trajectory point, the terminal performs lane recognition on the vehicle-mounted image corresponding to the target historical trajectory point to obtain a lane recognition result, and performs positioning on the GPS map based on the lane recognition result to obtain the actual matching lane of the target historical trajectory point. For each target historical trajectory point, the terminal compares the predicted matching lane with the actual matching lane to obtain the collection accuracy of the target historical trajectory point. For the target historical trajectory point set, the terminal calculates the collection accuracy of the target historical trajectory point set based on the collection accuracy of each target historical trajectory point. Specifically, for each target historical trajectory point, the terminal compares the predicted matching lane with the actual matching lane, and if the predicted matching lane is the same as the actual matching lane, the collection accuracy of the target historical trajectory point is set to 1, otherwise the collection accuracy of the target historical trajectory point is set to 0. For the target historical trajectory point set, the terminal calculates the average value (for the sake of convenience, referred to as average collection accuracy) of the collection accuracy of each target historical trajectory point, and takes the average collection accuracy as the collection accuracy of the target historical trajectory point set.

[0091] In a case where the collection accuracy of the target historical trajectory point set does not satisfy the preset collection accuracy condition, the terminal re-determines the target historical trajectory point in the historical trajectory point set based on the preset second selection strategy, and returns to execute the step of obtaining the target historical trajectory point set until the collection accuracy of the target historical trajectory point set satisfies the preset collection accuracy condition. Specifically, in a case where the collection accuracy of the target historical trajectory point set does not satisfy the preset collection accuracy condition, the terminal re-obtains the second preset number of initial historical trajectory points, and constructs a set by taking the second preset number of initial historical trajectory points as target historical trajectory points to obtain the target historical trajectory point set, and returns to execute the step of obtaining the target historical trajectory point set until the collection accuracy of the target historical trajectory point set satisfies the preset collection accuracy condition. It can be understood that the new target historical trajectory point set includes at least one historical trajectory point that is not in the original target historical trajectory point set. In a case where the collection accuracy of the target historical trajectory point set satisfies the preset collection accuracy condition, the terminal selects the target historical trajectory point as the correction reference point according to the collection accuracy of the target historical trajectory point for each target historical trajectory point in the target historical trajectory point set to obtain the correction reference point set.

[0092] In the embodiment of the application, in a case where the collection accuracy of the target historical trajectory point set does not satisfy the preset collection accuracy condition, the terminal re-determines the target historical trajectory point in the historical trajectory point set based on the preset second selection strategy, and returns to execute the step of obtaining the target historical trajectory point set until the collection accuracy of the target historical trajectory point set satisfies the preset collection accuracy condition. Specifically, in a case where the collection accuracy of the target historical trajectory point set does not satisfy the preset collection accuracy condition, the terminal re-obtains the second preset number of initial historical trajectory points, and constructs a set by taking the second preset number of initial historical trajectory points as target historical trajectory points to obtain the target historical trajectory point set, and returns to execute the step of obtaining the target historical trajectory point set until the collection accuracy of the target historical trajectory point set satisfies the preset collection accuracy condition. It can be understood that the new target historical trajectory point set includes at least one historical trajectory point that is not in the original target historical trajectory point set. In a case where the collection accuracy of the target historical trajectory point set satisfies the preset collection accuracy condition, the terminal selects the target historical trajectory point as the correction reference point according to the collection accuracy of the target historical trajectory point for each target historical trajectory point in the target historical trajectory point set to obtain the correction reference point set.

[0093] In the embodiment, the terminal determines the target historical trajectory point set according to the preset second selection strategy, and determines the correction reference point set according to the target historical trajectory point set in a case where the collection accuracy of the target historical trajectory point set satisfies the preset collection accuracy condition. Since the correction reference point set is obtained based on the target historical trajectory point set whose collection accuracy satisfies the preset collection accuracy condition, the correction quality of the correction reference point set is ensured, and the correction accuracy of trajectory correction based on the correction reference point set is improved.

[0094] In one embodiment, as shown in FIG. 4, the step of determining the collection trajectory point first subset in the collection trajectory point set according to the preset correction selection condition includes: Figure 4

[0095] In step 402, in a case where the arrival time of the collection trajectory point is the same as the shooting time of the vehicle-mounted image, the collection trajectory point is taken as the diversity trajectory point, and the arrival time of the diversity trajectory point is taken as the diversity time.

[0096] ​In the embodiments of the present application, for the collection trajectory point set, the terminal acquires the arrival time of the collection trajectory point, in the case that the arrival time of the collection trajectory point is the same as the shooting time of the vehicle-mounted image, the terminal takes the collection trajectory point as the diversity trajectory point, takes the arrival time of the diversity trajectory point as the diversity time, and obtains the diversity time set. The terminal takes the earliest diversity time in the diversity time set as the first diversity time. For example, assuming that the collection trajectory point set is {x1, x2, x3, x4, x5, x6}, the arrival time of the collection trajectory point x n is t n , where n is a positive integer and is less than or equal to 6. If the shooting time of the vehicle-mounted image P1 is t3 and the shooting time of the vehicle-mounted image P2 is t5, the terminal identifies that the arrival time t3 of the collection trajectory point x3 is the same as the shooting time t3 of the vehicle-mounted image P1, and takes the collection trajectory point x3 as the diversity trajectory point x3 and takes the arrival time t3 of the diversity trajectory point x3 as the diversity time t3. Similarly, the diversity trajectory point x5 and the diversity time t5 are obtained, and the terminal obtains the diversity time set {t3, t5}. The terminal compares the diversity time t3 and the diversity time t5, and obtains that the diversity time t3 is the earliest diversity time in the diversity time set {t3, t5}, and thus the terminal takes the diversity time t3 as the first diversity time.

[0097] In step 404, in the collection trajectory point set, the collection trajectory point with the arrival time earlier than or equal to the first diversity time is determined, and a first subset of collection trajectory points is obtained.

[0098] The first diversity time is the earliest diversity time.

[0099] In the embodiments of the present application, in the collection trajectory point set, the terminal acquires the collection trajectory point with the arrival time earlier than or equal to the first diversity time, and takes the collection trajectory point with the arrival time earlier than or equal to the first diversity time as an element of the first subset of collection trajectory points, and obtains the first subset of collection trajectory points. For example, assuming that the collection trajectory point set is {x1, x2, x3, x4, x5, x6}, the collection trajectory point set {x1, x2, x3, x4, x5, x6} is an ordered set, from left to right, the arrival time of the collection trajectory point is from early to late, the arrival time of the collection trajectory point x n is t n , where n is a positive integer and is less than or equal to 6, and the diversity time t3 is the first diversity time. In the collection trajectory point set {x1, x2, x3, x4, x5, x6}, the terminal acquires the sampling trajectory points x1, x2, and x3 with the arrival time earlier than or equal to the first diversity time t3, and takes x1, x2, and x3 as the elements of the first subset of collection trajectory points, and obtains the first subset of collection trajectory points {x1, x2, x3}.

[0100] In this embodiment, the terminal determines a first subset of the collected trajectory points according to a preset deviation correction selection condition. In this way, a premise of subset deviation correction is provided for subsequent processing.

[0101] In one embodiment, the deviation correction accuracy of the first subset of the initial deviation correction trajectory points is determined according to the first subset of the initial deviation correction trajectory points and the subset of the actual matching lane data corresponding to the first subset of the initial deviation correction trajectory points, and includes:

[0102] According to a preset third selection strategy, a target deviation correction trajectory point is determined in the first subset of the initial deviation correction trajectory points, and deviation correction trajectory data of the target deviation correction trajectory point is obtained. Lane matching is performed based on the deviation correction trajectory data of the target deviation correction trajectory point and lane data of the lane center line to determine a predicted matching lane of the target deviation correction trajectory point. The deviation correction accuracy of the target deviation correction trajectory point is determined based on the predicted matching lane of the target deviation correction trajectory point and the actual matching lane of the target deviation correction trajectory point, and the deviation correction accuracy of the first subset of the initial deviation correction trajectory points is determined according to the deviation correction accuracy of each target deviation correction trajectory point.

[0103] In this embodiment, the terminal selects a third preset number of initial deviation correction trajectory points as target deviation correction trajectory points in the first subset of the initial deviation correction trajectory points according to a preset third selection strategy. In one embodiment, the third preset number is 2. In the first subset of the initial deviation correction trajectory points, the terminal selects an initial deviation correction trajectory point corresponding to the earliest arrival time as a target deviation correction trajectory point, and selects an initial deviation correction trajectory point corresponding to the latest arrival time as another target deviation correction trajectory point. The terminal obtains deviation correction trajectory data of the target deviation correction trajectory point, and performs lane matching based on the deviation correction trajectory data of the target deviation correction trajectory point and lane data of the lane center line to obtain a predicted matching lane of the target deviation correction trajectory point. For each target deviation correction trajectory point, the terminal compares the predicted matching lane with the actual matching lane to obtain the deviation correction accuracy of the target deviation correction trajectory point. For the first subset of the initial deviation correction trajectory points, the terminal calculates the deviation correction accuracy of the first subset of the initial deviation correction trajectory points based on the deviation correction accuracy of each target deviation correction trajectory point. Specifically, for each target deviation correction trajectory point, the terminal compares the predicted matching lane with the actual matching lane. If the predicted matching lane is the same as the actual matching lane, the deviation correction accuracy of the target deviation correction trajectory point is set to 1, otherwise the deviation correction accuracy of the target deviation correction trajectory point is set to 0. For the first subset of the initial deviation correction trajectory points, the terminal calculates an average value of the deviation correction accuracy of each target deviation correction trajectory point (for convenience, referred to as average deviation correction accuracy), and takes the average deviation correction accuracy as the deviation correction accuracy of the first subset of the initial deviation correction trajectory points.

[0104] In this embodiment, the terminal determines the correction accuracy of the first subset of initial correction trajectory points according to the first subset of initial correction trajectory points and the subset of actual matching lane data corresponding to the first subset of initial correction trajectory points. This provides a prerequisite for the terminal to subsequently determine whether the correction accuracy of the first subset of initial correction trajectory points meets the preset correction accuracy condition.

[0105] In one embodiment, determining the set of correction reference points according to the set of target historical trajectory points comprises:

[0106] For each target historical trajectory point in the set of target historical trajectory points, if the collection accuracy of the target historical trajectory point meets the preset collection accuracy condition, the target historical trajectory point is taken as a correction reference point.

[0107] In the embodiments of this application, for each target historical trajectory point in the set of target historical trajectory points, the terminal acquires the collection accuracy of the target historical trajectory point. If the collection accuracy of the target historical trajectory point meets the preset collection accuracy condition, the terminal takes the target historical trajectory point as a correction reference point and constructs a set of correction reference points based on the correction reference points. In one embodiment, for each target historical trajectory point in the set of target historical trajectory points, the terminal takes the target correction trajectory point with a correction accuracy of 1 as a correction reference point.

[0108] In this embodiment, for each target historical trajectory point in the set of target historical trajectory points, the terminal takes the target historical trajectory point with a collection accuracy meeting the preset collection accuracy condition as a correction reference point. Therefore, the collection accuracy of the correction reference point is ensured, and the correction accuracy of trajectory correction based on the correction reference point is improved.

[0109] In one embodiment, after taking the first subset of initial correction trajectory points as the first subset of corrected correction trajectory points, the method further comprises:

[0110] The first subset of corrected correction trajectory points is taken as a new set of correction reference points, and the step of determining a target correction reference point in the set of correction reference points according to the preset first selection strategy is performed.

[0111] In the embodiment, the terminal takes the first subset of the corrected trajectory points as a new set of correction reference points, and returns to step 102 until all the collected trajectory points in the collected trajectory point set are corrected, and the terminal obtains at least one subset of the corrected trajectory points. The set of the corrected trajectory points includes a plurality of subsets of the corrected trajectory points, and each subset of the corrected trajectory points can be named as an Nth subset of the corrected trajectory points for convenience, where N is an integer and N is greater than or equal to 1. That is, the terminal takes the Nth subset of the corrected trajectory points as a new set of correction reference points, and returns to step 102 to obtain an (N+1)th subset of the corrected trajectory points. N is an integer and N is greater than or equal to 1. Assuming that the collected trajectory point set has M subsets, after (M-1) times of the above correction, the terminal obtains M subsets of the corrected trajectory points, where M is an integer and M is greater than or equal to 2. Specifically, assuming that the collected trajectory point set is {x1, x2, x3, x4, x5, x6}, the first subset of the sampling trajectory points is {x1, x2, x3}, the second subset of the sampling trajectory points is {x4, x5}, and the third subset of the sampling trajectory points is {x6}. The first subset of the corrected trajectory points is {x′1, x′2, x′3}. The terminal takes the first subset of the corrected trajectory points as a new set of correction reference points, and returns to step 102 to obtain the second subset of the corrected trajectory points corresponding to the second subset of the collected trajectory points. The terminal takes the second subset of the corrected trajectory points as a new set of correction reference points, and returns to step 102 to obtain the third subset of the corrected trajectory points corresponding to the third subset of the collected trajectory points. The terminal stops correction and obtains three subsets of the corrected trajectory points {x′1, x′2, x′3}, {x′4, x′5}, and {x′6}.

[0112] In the embodiment, the terminal corrects the (N+1)th subset of the collected trajectory points based on the Nth subset of the corrected trajectory points as a new set of correction reference points to obtain an initial (N+1)th subset of the corrected trajectory points. When the correction accuracy of the initial (N+1)th subset of the corrected trajectory points meets the preset correction accuracy, the initial (N+1)th subset of the corrected trajectory points is taken as the (N+1)th subset of the corrected trajectory points. Therefore, the correction accuracy of each subset of the corrected trajectory points meets the preset correction accuracy, and the correction accuracy of trajectory correction is improved.

[0113] In one embodiment, according to a preset first selection strategy, the target correction reference point is determined from the set of correction reference points, including:

[0114] From the set of correction reference points, select the correction reference point corresponding to the latest arrival time as the target correction reference point.

[0115] In this embodiment, for each correction reference point in the correction reference point set, the terminal compares the arrival times of each correction reference point and selects the correction reference point with the latest arrival time as the target correction reference point. Regarding step 110, if the correction accuracy of the first subset of the initial correction trajectory points does not meet the preset correction accuracy condition, for each correction reference point in the correction reference point set other than the already selected correction reference points, the terminal selects the correction reference point with the latest arrival time as the new target correction reference point. Here, the already selected correction reference point refers to a correction reference point that has been selected by the terminal as a target correction reference point. That is, the terminal will not repeatedly select the same correction reference point as a target correction reference point. Specifically, assuming the correction reference point set is {z1, z2, z3, ..., z...} (n-1) , z n}, and the set of correction reference points is an ordered set, from left to right, with the arrival times of the correction reference points from earliest to latest, where n is a positive integer. If the correction reference point z n Given selected correction reference points, if the correction accuracy of the first subset of the initial correction trajectory points does not meet the preset correction accuracy condition, then for each correction reference point z1, z2, z3, ..., z in the correction reference point set... (n-2) , z (n-1) The terminal will use the latest arrival time as the correction reference point z. (n-1) As a new reference point for correcting deviations from the target.

[0116] In this embodiment, the terminal selects the correction reference point corresponding to the latest arrival time as the target correction reference point. In this way, the arrival time of the target correction reference point is closer to the arrival time of the collected trajectory points. Therefore, the correction processing of the collected trajectory points is performed based on the target correction reference point with a closer arrival time, resulting in higher correction accuracy.

[0117] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0118] Based on the same inventive concept, the embodiments of the present application also provide a trajectory correction device for implementing the trajectory correction method described above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more trajectory correction device embodiments provided below can refer to the limitations of the trajectory correction method described above, which will not be described here.

[0119] In one embodiment, as shown in Figure 5 A trajectory correction device is provided, comprising:

[0120] The first determination module 502 is configured to determine a target correction reference point in the set of correction reference points according to a preset first selection strategy.

[0121] The second determination module 504 is configured to determine a first subset of collected trajectory points in the set of collected trajectory points according to a preset correction selection condition.

[0122] The third determination module 506 is configured to determine a first subset of initial correction trajectory points corresponding to the first subset of collected trajectory points based on the trajectory data of the target correction reference point, a subset of arrival times corresponding to the first subset of collected trajectory points, and a pre-trained position correction model.

[0123] The fourth determination module 508 is configured to determine the correction accuracy of the first subset of initial correction trajectory points according to the first subset of initial correction trajectory points and a subset of actual matching lane data corresponding to the first subset of initial correction trajectory points.

[0124] The fifth determination module 510 is configured to re-determine the target correction reference point in the set of correction reference points based on the preset first selection strategy if the correction accuracy does not meet the preset correction accuracy condition, and return to perform the step of determining the first subset of collected trajectory points according to the preset correction selection condition until the correction accuracy meets the preset accuracy condition, and the first subset of initial correction trajectory points is taken as the first subset of corrected correction trajectory points.

[0125] In one embodiment, the trajectory correction device further comprises:

[0126] The sixth determination module is configured to determine a target historical trajectory point in the set of historical trajectory points according to a preset second selection strategy to obtain a set of target historical trajectory points.

[0127] The seventh determination module is configured to perform lane matching based on the trajectory data of the target historical trajectory point and the lane data of the lane center line to determine the predicted matching lane of the target historical trajectory point.

[0128] The eighth determining module is configured to determine the collection accuracy of the trajectory data of the target historical trajectory point based on the predicted matching lane of the target historical trajectory point and the actual matching lane of the target historical trajectory point, and determine the collection accuracy of the target historical trajectory point set according to the collection accuracy of each target historical trajectory point.

[0129] The ninth determining module is configured to, in a case where the collection accuracy of the target historical trajectory point set does not satisfy the preset collection accuracy condition, determine the target historical trajectory point in the historical trajectory point set again based on a preset second selection strategy, and return to execute the step of obtaining the target historical trajectory point set until the collection accuracy of the target historical trajectory point set satisfies the preset collection accuracy condition, and determine the set of correction reference points based on the target historical trajectory point set.

[0130] In an embodiment, the second determining module 504 is specifically configured to:

[0131] In a case where the arrival time of the collection trajectory point is the same as the shooting time of the vehicle-mounted image, the collection trajectory point is taken as a diversity trajectory point, and the arrival time of the diversity trajectory point is taken as a diversity time.

[0132] In the collection trajectory point set, a collection trajectory point with an arrival time earlier than or equal to a first diversity time is determined to obtain a first subset of collection trajectory points. The first diversity time is the earliest diversity time.

[0133] In an embodiment, the fourth determining module 508 is specifically configured to:

[0134] According to a preset third selection strategy, a target correction trajectory point is determined in the first subset of initial correction trajectory points, and correction trajectory data of the target correction trajectory point is obtained.

[0135] Lane matching is performed based on the correction trajectory data of the target correction trajectory point and lane data of the lane center line to determine a predicted matching lane of the target correction trajectory point.

[0136] The correction accuracy of the target correction trajectory point is determined based on the predicted matching lane of the target correction trajectory point and the actual matching lane of the target correction trajectory point, and the correction accuracy of the first subset of initial correction trajectory points is determined according to the correction accuracy of each target correction trajectory point.

[0137] In an embodiment, the ninth determining module is specifically configured to:

[0138] For each target historical trajectory point in the target historical trajectory point set, the target historical trajectory point is taken as a correction reference point in a case where the collection accuracy of the target historical trajectory point satisfies the preset collection accuracy condition.

[0139] In an embodiment, the trajectory correction device further comprises:

[0140] The circulation module is configured to take the first subset of the corrected trajectory points as a new set of correction reference points, and return to performing the step of determining a target correction reference point from the set of correction reference points according to a preset first selection strategy.

[0141] In one embodiment, the first determination module 502 is specifically configured to:

[0142] Select a correction reference point corresponding to the latest arrival time from the set of correction reference points as the target correction reference point.

[0143] The above-mentioned modules in the trajectory correction device can be realized by software, hardware, and combinations thereof, in whole or in part. The above-mentioned modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so as to be called and executed by the processor to perform the operations corresponding to the above-mentioned modules.

[0144] In one embodiment, a computer device is provided, which can be a terminal, and the internal structure diagram thereof can be as shown in Figure 6 The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The communication interface of the computer device is configured to perform wired or wireless communication with an external terminal. The wireless communication can be achieved through WIFI, mobile cellular network, NFC (Near Field Communication), or other technologies. The computer program is executed by the processor to implement a trajectory correction method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or can be a key, trackball, or touchpad arranged on the shell of the computer device. In addition, an external keyboard, touchpad, or mouse, etc. can also be used.

[0145] Those skilled in the art can understand that Figure 6 the structure shown in the above-mentioned figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. Specifically, the computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0146] In an embodiment, a computer device is also provided, including a memory and a processor, the memory storing a computer program, and the processor implementing the steps in the above method embodiments when executing the computer program.

[0147] In an embodiment, a computer readable storage medium is provided, storing a computer program, and the computer program implementing the steps in the above method embodiments when executed by a processor.

[0148] In an embodiment, a computer program product is provided, including a computer program, and the computer program implementing the steps in the above method embodiments when executed by a processor.

[0149] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties.

[0150] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (Read-Only Memory, ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric memory (Ferroelectric Random Access Memory, FRAM), phase change memory (Phase Change Memory, PCM), graphene memory, etc. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0151] Any combination of the technical features of the above embodiments can be made. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0152] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent of the present application. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of protection of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A trajectory deviation correction method, characterized by, The method comprises: determining a target rectification reference point in a rectification reference point set according to a preset first selection strategy; determining a first subset of collected trajectory points in the collected trajectory point set according to a preset rectification selection condition; determining a first subset of initial rectified trajectory points corresponding to the first subset of collected trajectory points based on trajectory data of the target rectification reference point, a subset of arrival times corresponding to the first subset of collected trajectory points, and a pre-trained position rectification model; the trajectory data of the target rectification reference point comprises a longitude of the target rectification reference point, a latitude of the target rectification reference point, an altitude of the terminal, a speed of the terminal, a collection time of a global positioning system device, and a direction of the terminal; the subset of arrival times corresponding to the first subset of collected trajectory points comprises an arrival time of at least one collected trajectory point; determining rectification accuracy of the first subset of initial rectified trajectory points according to the first subset of initial rectified trajectory points and a subset of actual matching lane data corresponding to the first subset of initial rectified trajectory points; in a case where the rectification accuracy does not satisfy a preset rectification accuracy condition, re-determining the target rectification reference point in the rectification reference point set based on the preset first selection strategy, and returning to perform the step of determining the target rectification reference point in the rectification reference point set according to the preset rectification selection condition until the rectification accuracy satisfies the preset accuracy condition, and taking the first subset of initial rectified trajectory points as a first subset of rectified trajectory points; the determining of the first subset of initial rectified trajectory points based on the trajectory data of the target rectification reference point, the subset of arrival times corresponding to the first subset of collected trajectory points, and the pre-trained position rectification model comprises: performing lane matching processing according to the trajectory data of the target rectification reference point and lane data of a lane center line to obtain a predicted matching lane of the target rectification reference point; performing rectification processing according to the trajectory data of the target rectification reference point and the predicted matching lane of the target rectification reference point to obtain reference trajectory data of the target rectification reference point; inputting the reference trajectory data of the target rectification reference point and the arrival times of the collected trajectory points in the first subset of collected trajectory points into the pre-trained position rectification model to output predicted longitude and latitude of the collected trajectory points; updating the predicted longitude and latitude of the collected trajectory points as longitude and latitude in the trajectory data of the collected trajectory points to obtain initial rectified trajectory data, and further obtain the first subset of initial rectified trajectory points.

2. The method of claim 1, wherein, before the determining of the target rectification reference point in the rectification reference point set according to the preset first selection strategy, the method further comprises: determining a target historical trajectory point in a historical trajectory point set according to a preset second selection strategy to obtain a target historical trajectory point set; determining a predicted matching lane of the target historical trajectory point based on trajectory data of the target historical trajectory point and lane data of a lane center line; Determine the collection accuracy of the target historical trajectory point based on the predicted matching lane of the target historical trajectory point and the actual matching lane of the target historical trajectory point, and determine the collection accuracy of the target historical trajectory point set based on the collection accuracy of each target historical trajectory point. In a case where the collection accuracy of the target historical trajectory point set does not satisfy the preset collection accuracy condition, determine the target historical trajectory point in the historical trajectory point set again based on the preset second selection strategy, and return to execute the step of obtaining the target historical trajectory point set until the collection accuracy of the target historical trajectory point set satisfies the preset collection accuracy condition, and determine the deviation reference point set based on the target historical trajectory point set.

3. The method of claim 1, wherein, The step of determining the first subset of collection trajectory points from the collection trajectory point set based on the preset deviation selection condition comprises: In a case where the arrival time of the collection trajectory point is the same as the shooting time of the vehicle-mounted image, the collection trajectory point is taken as a diversity trajectory point, and the arrival time of the diversity trajectory point is taken as a diversity time. Determine the collection trajectory point in the collection trajectory point set whose arrival time is earlier than or equal to the first diversity time to obtain the first subset of collection trajectory points, and the first diversity time is the earliest diversity time.

4. The method according to any one of claims 1 to 3, characterized in that, The step of determining the deviation accuracy of the first subset of initial deviation trajectory points based on the first subset of initial deviation trajectory points and the actual matching lane data subset corresponding to the first subset of initial deviation trajectory points comprises: Determine a target deviation trajectory point in the first subset of initial deviation trajectory points based on a preset third selection strategy, and obtain the deviation trajectory data of the target deviation trajectory point. Perform lane matching based on the deviation trajectory data of the target deviation trajectory point and the lane data of the lane center line to determine the predicted matching lane of the target deviation trajectory point. Determine the deviation accuracy of the target deviation trajectory point based on the predicted matching lane of the target deviation trajectory point and the actual matching lane of the target deviation trajectory point, and determine the deviation accuracy of the first subset of initial deviation trajectory points based on the deviation accuracy of each target deviation trajectory point.

5. The method of claim 2, wherein, The step of determining the deviation reference point set based on the target historical trajectory point set comprises: In a case where the collection accuracy of the target historical trajectory point satisfies the preset collection accuracy condition, take the target historical trajectory point as the deviation reference point.

6. The method of claim 1, wherein, After taking the first subset of initial deviation trajectory points as the first subset of deviation trajectory points after deviation, further comprise: Take the first subset of deviation trajectory points after deviation as a new deviation reference point set, and return to execute the step of determining the target deviation reference point in the deviation reference point set based on the preset first selection strategy.

7. The method of claim 1, wherein, The step of determining the target deviation reference point in the deviation reference point set based on the preset first selection strategy comprises: Select the deviation reference point corresponding to the latest arrival time from the deviation reference point set as the target deviation reference point.

8. A trajectory deviation rectifying device, characterized by, The device comprises: A first determining module configured to determine a target rectification reference point from a set of rectification reference points according to a preset first selection strategy; A second determining module configured to determine a first subset of collected trajectory points from a set of collected trajectory points according to a preset rectification selection condition; A third determining module configured to determine a first subset of initial rectified trajectory points corresponding to the first subset of collected trajectory points based on trajectory data of the target rectification reference point, a subset of arrival times corresponding to the first subset of collected trajectory points, and a pre-trained position rectification model; the trajectory data of the target rectification reference point comprises a longitude of the target rectification reference point, a latitude of the target rectification reference point, an altitude of a terminal, a speed of the terminal, a collection time of a global positioning system device, and a direction of the terminal; the subset of arrival times corresponding to the first subset of collected trajectory points comprises arrival times of at least one collected trajectory point; A fourth determining module configured to determine rectification accuracy of the first subset of initial rectified trajectory points according to the first subset of initial rectified trajectory points and a subset of actual matching lane data corresponding to the first subset of initial rectified trajectory points; A fifth determining module configured to, in a case where the rectification accuracy does not satisfy a preset rectification accuracy condition, re-determine the target rectification reference point from the set of rectification reference points based on the preset first selection strategy, and return to perform the step of determining according to the preset rectification selection condition until the rectification accuracy satisfies the preset accuracy condition, and take the first subset of initial rectified trajectory points as a first subset of rectified trajectory points; The third determining module is specifically configured to perform lane matching processing according to the trajectory data of the target rectification reference point and lane data of a lane center line to obtain a predicted matching lane of the target rectification reference point, perform rectification processing according to the trajectory data of the target rectification reference point and the predicted matching lane of the target rectification reference point to obtain reference trajectory data of the target rectification reference point, and input the reference trajectory data of the target rectification reference point and arrival times of collected trajectory points in the first subset of collected trajectory points into the pre-trained position rectification model to output predicted longitude and latitude of the collected trajectory points; and the predicted longitude and latitude of the collected trajectory points are updated as longitude and latitude in the trajectory data of the collected trajectory points to obtain initial rectified trajectory data, and further obtain the first subset of initial rectified trajectory points. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. The processor executes the computer program to implement the steps of the method in any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 7.

Images