Driving track restoration method and device based on multi-source data

Abstract

The invention discloses a driving track restoration method and device based on multi-source data. The method comprises the steps that a GPS track point set, a dead reckoning point set and road network data are acquired; for each GPS track point, determining a plurality of reference points on a plurality of road sections; for each reference point, calculating a calculation position point by combining the direction and the speed from the GPS track point to the next GPS track point; calculating to obtain a new reference point by taking the reference point as a starting point and combining the driving direction of the road section where the reference point is located and the time difference and the speed from the GPS track point to the next GPS track point; taking the new reference point as a starting point, repeating the steps of calculating and updating the reckoning position point until the next GPS track point does not exist, and obtaining a reckoning position point set; and based on the road network data, calculating a matching probability and an evolution probability associated with each to-be-matched track point in the GPS track point set, the dead reckoning point set and the reckoning position point set, and obtaining a track reduction result. According to the method, the accuracy and robustness of trajectory reduction are effectively improved.

Description

Technical Field

[0001] The invention relates to a method and device for restoring a vehicle's driving trajectory based on multi-source data. Background Art

[0002] Trajectory restoration is an indispensable technology in the field of transportation and intelligent vehicles, which aims to accurately match the vehicle's GPS location information with the road network. This process is crucial because it not only improves the accuracy of navigation systems, but also supports intelligent traffic management, autonomous driving, and travel data analysis.

[0003] By restoring vehicle trajectories, transportation systems can better understand and predict traffic flows, improve road utilization efficiency, and provide important data support for accident prevention and traffic planning. Summary of the invention

[0004] In order to obtain a more accurate trajectory restoration result, an embodiment of the present invention provides a method and device for restoring a vehicle trajectory based on multi-source data.

[0005] In a first aspect, an embodiment of the present invention provides a method for restoring a vehicle trajectory based on multi-source data, which may include:

[0006] Acquire a GPS track point set, a dead reckoning point set and a road network data; the road network data includes a plurality of road sections; the GPS track point set includes a plurality of GPS track points sorted in time from earliest to latest;

[0007] For each GPS track point, multiple reference points are determined on multiple road sections;

[0008] For each reference point, taking the reference point as the starting point, a calculated position point is calculated by combining the direction and speed from the GPS track point to the next GPS track point;

[0009] Taking the reference point as the starting point, a new reference point is calculated based on the driving direction of the road section where the reference point is located, the time difference from the GPS track point to the next GPS track point, and the speed;

[0010] Taking the new reference point as the starting point, repeat the above steps of calculating the estimated position point and updating the reference point until there is no next GPS track point, and combine all the estimated position points to obtain an estimated position point set;

[0011] Based on the road network data, the matching probability and the evolution probability associated with each to-be-matched trajectory point in the GPS trajectory point set, the dead reckoning point set and the estimated position point set are calculated to obtain a trajectory restoration result.

[0012] In one or some optional implementations of the embodiment of the present application, for each GPS track point, determining multiple reference points on multiple road sections includes:

[0013] For each GPS track point, based on the position of each road segment in the road network data, determine a plurality of candidate road segments corresponding to the GPS track point;

[0014] The GPS track points are projected onto each candidate road section to obtain multiple reference points.

[0015] In one or some optional implementations of the embodiment of the present application, for each reference point, taking the reference point as a starting point, and combining the direction and speed from the GPS track point to the next GPS track point, a calculated position point is calculated, including:

[0016] Calculate the estimated distance based on the time difference and speed from the GPS track point to the next GPS track point;

[0017] For each reference point, the reference point is used as a starting point, and the latitude and longitude coordinates of the estimated position point are calculated according to the latitude and longitude coordinates of the reference point, the direction and the estimated distance.

[0018] In one or some optional implementations of the embodiments of the present application, the method of calculating the matching probability and the evolution probability associated with each to-be-matched trajectory point in the GPS trajectory point set, the dead reckoning point set, and the estimated position point set based on the road network data to obtain a trajectory restoration result includes:

[0019] Based on the position of each road section in the road network data, determining a plurality of candidate road sections corresponding to each to-be-matched trajectory point;

[0020] Combining each to-be-matched trajectory point with each corresponding candidate road segment into a candidate matching pair;

[0021] Calculating the matching probability and associated multiple evolution probabilities corresponding to each candidate matching pair according to the distance between the to-be-matched trajectory point and the candidate road section in the candidate matching pair;

[0022] Constructing multiple candidate point sequences according to the timestamps of the to-be-matched trajectory points in each candidate matching pair;

[0023] Calculate the probability of each candidate point sequence according to the matching probability corresponding to each candidate matching pair and the associated multiple evolution probabilities;

[0024] The candidate point sequence with the highest probability is selected to obtain the trajectory restoration result.

[0025] In one or some optional implementations of the embodiment of the present application, after obtaining multiple candidate point sequences, the following step further includes:

[0026] In a candidate point sequence, if the distance between the to-be-matched trajectory points in two adjacent candidate matching pairs is greater than the product of the time difference between the two to-be-matched trajectory points and the preset maximum driving speed, then the candidate point sequence is deleted; and / or

[0027] In a candidate point sequence, if there is no topological relationship between the candidate road segments in two adjacent candidate matching pairs, the candidate point sequence is deleted.

[0028] In one or some optional implementations of the embodiment of the present application, the calculation of the matching probability corresponding to each candidate matching pair and the associated multiple evolution probabilities according to the distance between the to-be-matched trajectory point and the candidate road section in the candidate matching pair includes:

[0029] Calculating the matching probability of each candidate matching pair according to the projection distance between the to-be-matched trajectory point and the candidate road section in the candidate matching pair;

[0030] According to the timestamps of the to-be-matched trajectory points in each candidate matching pair, all candidate matching pairs are divided into a plurality of time point candidate sets sorted in order of time from earliest to latest;

[0031] For each candidate matching pair in the candidate set at each time point, the evolution probability between the candidate matching pair and all candidate matching pairs in the candidate set at the next time point is calculated.

[0032] In one or some optional implementations of the embodiment of the present application, multiple candidate point sequences are constructed according to the timestamps of the trajectory points to be matched in each candidate matching pair, including:

[0033] Take any candidate matching pair from each time point candidate set, construct a candidate point sequence in order, and repeat the operation of constructing the candidate point sequence until all time point candidate sets are traversed to obtain all possible candidate point sequences, thereby obtaining the multiple candidate point sequences.

[0034] In one or some optional implementations of the embodiment of the present application, the probability of each candidate point sequence is calculated according to the matching probability corresponding to each candidate matching pair and the associated multiple evolution probabilities, including:

[0035] The probability of each candidate point sequence is calculated according to the matching probability of each candidate matching pair in the candidate point sequence and the evolution probability between each two adjacent candidate matching pairs.

[0036] In one or some optional implementations of the embodiment of the present application, after all candidate matching pairs are divided into a plurality of time point candidate sets sorted in order from earliest to latest, the following further includes:

[0037] For each candidate matching pair in the candidate set at each time point, connect the candidate matching pair with each candidate matching pair in the candidate set at the previous time point to obtain multiple ternary permutations corresponding to the candidate matching pair;

[0038] Calculating the probability of each ternary arrangement according to the matching probability of each candidate matching pair in the ternary arrangement and the evolution probability between adjacent candidate matching pairs;

[0039] Selecting the maximum value of the probabilities among all the ternary arrangements corresponding to the candidate matching pair as the probability of the candidate matching pair;

[0040] According to the probabilities of all candidate matching pairs in the candidate set at each time point, a preset candidate threshold number of candidate matching pairs with the highest probabilities in the candidate set at the time point is retained.

[0041] In a second aspect, an embodiment of the present invention provides a vehicle trajectory restoration device based on multi-source data, which may include:

[0042] A first acquisition module is used to acquire a GPS track point set, a dead reckoning point set and a road network data; the road network data includes a plurality of road sections; the GPS track point set includes a plurality of GPS track points sorted in time from earliest to latest;

[0043] A reference point determination module is used to determine multiple reference points on multiple road sections for each GPS track point;

[0044] The estimated position module is used to calculate a estimated position point for each reference point, taking the reference point as a starting point and combining the direction and speed from the GPS track point to the next GPS track point;

[0045] The reference point update module is used to calculate a new reference point based on the reference point as the starting point, the driving direction of the road section where the reference point is located, the time difference from the GPS track point to the next GPS track point, and the speed;

[0046] A first combining module is used to take the new reference point as a starting point, repeat the steps of calculating the estimated position point and updating the reference point by the estimated position module and the updating reference point module, until there is no next GPS track point, and combine all the estimated position points to obtain an estimated position point set;

[0047] The first calculation module is used to calculate the matching probability and evolution probability associated with each to-be-matched trajectory point in the GPS trajectory point set, the dead reckoning point set and the estimated position point set based on the road network data to obtain a trajectory restoration result.

[0048] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program / instruction stored thereon, which, when executed by a processor, implements the above-mentioned method for restoring a vehicle trajectory based on multi-source data.

[0049] In a fourth aspect, an embodiment of the present invention provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the above-mentioned method for restoring a vehicle's driving trajectory based on multi-source data.

[0050] In a fifth aspect, an embodiment of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory, and when the processor executes the computer program, the method for restoring a vehicle's driving trajectory based on multi-source data as described above is implemented.

[0051] The beneficial effects of the above technical solution provided by the embodiment of the present invention include at least:

[0052] The embodiment of the present invention provides a driving trajectory restoration method based on multi-source data. The method obtains GPS trajectory point set, dead reckoning point set and road network data, and based on each GPS trajectory point and the direction and speed information of each GPS trajectory point, cyclically calculates to obtain the estimated position point set, and then calculates the matching probability and evolution probability associated with each to-be-matched trajectory point in the GPS trajectory point set, the dead reckoning point set and the estimated position point set based on the road network data, to obtain the trajectory restoration result. The method combines the position data of multiple sources and integrates them to perform trajectory restoration. Even if a certain data source has a problem, other data sources can still provide supplementary information to ensure that the quality of the overall trajectory restoration is not affected by the failure of a single data source, and effectively solves the problem of trajectory restoration errors caused by the lack or inaccuracy of GPS signals in traditional trajectory restoration methods in complex environments (such as tunnels, under viaducts, urban areas with tall buildings, etc.), which not only improves the accuracy of trajectory restoration, but also enhances robustness and adaptability, and provides more reliable technical support for the fields of intelligent transportation systems, vehicle navigation services, etc.

[0053] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0054] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0056] Figure 1 A schematic diagram of a flow chart of a vehicle trajectory restoration method based on multi-source data provided by an embodiment of the present invention;

[0057] Figure 2 A schematic diagram of the correct trajectory restoration effect of the prior art provided by an embodiment of the present invention;

[0058] Figure 3 A schematic diagram of multi-source trajectory data and road network data provided by an embodiment of the present invention;

[0059] Figure 4 A schematic diagram of the error effect of trajectory restoration in the prior art provided by an embodiment of the present invention;

[0060] Figure 5 A schematic diagram of a method for calculating a position point provided by an embodiment of the present invention;

[0061] Figure 6 A schematic diagram of the effect of realizing correct trajectory restoration based on estimated position points provided by an embodiment of the present invention;

[0062] Figure 7 An example diagram of vehicle trajectory restoration based on multi-source data provided by an embodiment of the present invention;

[0063] Figure 8 A schematic diagram of a matching probability distribution curve provided by an embodiment of the present invention;

[0064] Fig. 9 An example diagram of matching probability provided by an embodiment of the present invention;

[0065] Fig.10 An example diagram of evolution probability provided by an embodiment of the present invention;

[0066] Fig.11 A schematic diagram of the probability of candidate matching pairs provided by an embodiment of the present invention;

[0067] Fig.12 An example diagram of screening candidate matching pairs according to probability provided by an embodiment of the present invention;

[0068] Fig.13 An example diagram of a candidate point sequence without topological relationship provided by an embodiment of the present invention;

[0069] Fig.14 A schematic diagram of the structure of a vehicle trajectory restoration device based on multi-source data provided in an embodiment of the present application. DETAILED DESCRIPTION

[0070] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0071] The inventors found that traditional solutions only conduct trajectory restoration research based on GPS points. When the GPS accuracy is high, the quality of restoration will also be high. However, in certain environments, such as tunnels, viaducts, buildings, and severe weather conditions, the GPS signal may have large deviations or be lost, thereby affecting the accuracy of restoration and causing restoration errors or even failure.

[0072] For example, Figure 2 As shown, L1 to L10 represent road sections in the road network data, G1 to G5 are GPS track points, and when the track restoration is performed based on the existing technology, the track restoration result can be obtained as L1+L2+L3+L4, indicating that the vehicle's driving route is L1-L2-L3-L4.

[0073] However, when L2 and L3 are in tunnels or under viaducts, the GPS track point positions may be abnormal or even without GPS signals. At this time, if the track restoration is still performed based on the existing technology, the track restoration result will be wrong, such as Figure 3 As shown, G1 and G5 are correct GPS positions, while G2, G3, and G4 are abnormal GPS track points, which lead to restoration errors or even failures.

[0074] Based on this, the inventors have made the present invention after further research and development, providing a method and device for restoring a vehicle's driving trajectory based on multi-source data.

[0075] Embodiment 1

[0076] In the first embodiment of the present invention, a method for restoring a vehicle trajectory based on multi-source data is provided. Figure 1 As shown, the method may include the following steps S101-S106:

[0077] S101: Obtaining a GPS track point set, a dead reckoning point set and a road network data. The road network data includes a plurality of road segments. The GPS track point set includes a plurality of GPS track points sorted in time from earliest to latest. The dead reckoning (DR) point set includes a plurality of dead reckoning points sorted in time from earliest to latest, wherein the dead reckoning points are obtained by performing position calculation by combining GPS and sensor information.

[0078] S102: For each GPS track point, determine multiple reference points on multiple road sections.

[0079] S103: For each reference point, a calculated position point is calculated based on the direction and speed from the GPS track point to the next GPS track point, taking the reference point as the starting point.

[0080] S104: Taking the reference point as the starting point, a new reference point is calculated based on the driving direction of the road section where the reference point is located, the time difference from the GPS track point to the next GPS track point, and the speed.

[0081] S105: Taking the new reference point as the starting point, repeat the above steps S103 and S104 until there is no next GPS track point, and combine all the estimated position points to obtain an estimated position point set.

[0082] S106: Based on the road network data, the matching probability and the evolution probability associated with each to-be-matched trajectory point in the GPS trajectory point set, the dead reckoning point set and the estimated position point set are calculated to obtain a trajectory restoration result.

[0083] The embodiment of the present invention provides a driving trajectory restoration method based on multi-source data. The method obtains a GPS trajectory point set, a dead reckoning point set and a road network data, and based on each GPS trajectory point and the direction and speed information of each GPS trajectory point, cyclically calculates to obtain a calculated position point set, and then calculates the matching probability and evolution probability associated with each to-be-matched trajectory point in the GPS trajectory point set, the dead reckoning point set and the calculated position point set based on the road network data, to obtain a trajectory restoration result. The method effectively solves the problem of trajectory restoration errors caused by the lack or inaccuracy of GPS signals in traditional trajectory restoration methods in complex environments (such as tunnels, under viaducts, urban areas with tall buildings, etc.), not only improves the accuracy of trajectory restoration, but also enhances robustness and adaptability, and provides more reliable technical support for the fields of intelligent transportation systems, vehicle navigation services, etc.

[0084] In order to facilitate those skilled in the art to understand the present solution, the GPS track point set, dead reckoning point set and road network data provided in step S101 of the present embodiment are described more clearly and completely. Figure 4 As shown, L1 to L5 are road segments in the road network data, G1 to G5 represent track points in the GPS track point set, and D1 to D5 represent dead reckoning points (DR) in the dead reckoning point set.

[0085] Among them, the dead reckoning point is the position calculated based on GPS information and some sensor information. The specific calculation method is current technology and will not be described in detail here. Generally speaking, the position calculated by DR is more accurate, but the dead reckoning point requires real-time calculation results, and when the vehicle does not follow the scheduled route, the calculated position will jump. In addition, in addition to collecting GPS location information, the vehicle can also collect vehicle status information. Vehicle status information includes speed, direction, and acceleration. These parameters rely on location information to a certain extent to achieve accurate positioning.

[0086] like Figure 4 As shown, the planned route of the vehicle is L1-L2-L3-L4, but the actual route of the vehicle is L1-L5. At this time, the dead reckoning points D2, D3, and D4 jump in position until D5 jumps back to the correct position.

[0087] It can be seen from this that if the dead reckoning points are used as the only data source for the trajectory restoration method, it will still lead to errors in the trajectory restoration.

[0088] In the above step S102, for each GPS track point, multiple reference points corresponding to the GPS track point are determined on multiple road sections. Specifically, the following steps S1021-S1022 are included:

[0089] S1021: For each GPS track point, based on the position of each road segment in the road network data, determine a plurality of candidate road segments corresponding to the GPS track point.

[0090] Specifically, a corresponding first preset proximity range may be determined based on the GPS track point, and if a road section in the road network data is within the preset proximity range, the road section is determined as a candidate road section corresponding to the GPS track point.

[0091] Among them, the first preset proximity range can be determined in the following two ways: the first way is to set the first preset radius with the GPS track point as the center of the circle, and the range of the circle is the first preset proximity range. The second way is to set the first preset rectangular side length with the GPS track point as the center, and determine the rectangular area along the longitude and latitude lines according to the first preset rectangular side length, and the rectangular area is the first preset proximity range. Of course, those skilled in the art can also choose the method for determining the first preset proximity range according to actual conditions, and no specific limitation is made in the embodiments of the present application.

[0092] S1022: Projecting the GPS track points onto each candidate road section to obtain multiple reference points.

[0093] In the above step S103, for each reference point, a calculated position point is calculated with the reference point as the starting point and the direction and speed from the GPS track point to the next GPS track point. Specifically, the following steps S1031-S1032 are included:

[0094] S1031: Calculate the estimated distance based on the time difference and speed from the GPS track point to the next GPS track point.

[0095] Specifically, the estimated distance may be obtained by multiplying the time difference and the speed from a GPS track point to the next GPS track point.

[0096] The speed represents the average speed from a GPS track point to the next GPS track point, which can be obtained by a vehicle speed sensor and will not be described in detail here.

[0097] S1032: For each reference point, taking the reference point as the starting point, the latitude and longitude coordinates of the estimated position point are calculated according to the latitude and longitude coordinates, direction and estimated distance of the reference point.

[0098] Specifically, the heading angle θ and the length d of the great circle arc may be determined according to the latitude and longitude coordinates, direction and estimated distance of the accurate point, and the latitude coordinates of the estimated position point may be obtained based on the following formula 1:

[0099] φ2=arcsin(sin(φ1)*cos(d)+cos(φ1)*sin(d)*cos(θ)) Formula 1

[0100] Where φ2 is the latitude of the estimated position point, φ1 is the latitude of the reference point, arcsin represents the inverse sine function, sin and cos represent the sine and cosine functions respectively.

[0101] According to the latitude and longitude coordinates, direction and estimated distance of the accurate point, the longitude coordinates of the estimated position point are calculated based on the following formula 2:

[0102] λ2=λ1+arcsin2(sin(θ)*sin(d)*cos(φ1),cos(d)-sin(φ1)*sin(φ2))Formula 2

[0103] Where λ2 is the longitude of the estimated position point, λ1 is the longitude of the reference point, and arctan2 represents the inverse tangent function.

[0104] In the above step S104, the reference point is taken as the starting point, and the driving direction of the road section where the reference point is located, the time difference from the GPS track point to the next GPS track point, and the speed are combined to calculate a new reference point. Specifically, based on the time difference and speed from the GPS track point to the next GPS track point, the time difference is multiplied by the speed to obtain an estimated distance M in meters. Then, taking the reference point as the starting point, the new reference point is obtained by deducing M meters along the driving direction of the road section where the reference point is located.

[0105] In the embodiment of the present application, a new reference point is determined in combination with the driving direction of the road section, rather than directly using the estimated position point as the new reference point, because during the vehicle's driving, turning or changing lanes may cause an angle deviation between the estimated position point and the actual driving route. If this deviation is not corrected in time, it will accumulate in subsequent projection calculations, resulting in increased positioning errors. In this method, by calculating a new projection point along the actual driving direction of the vehicle, the moving trajectory can be effectively tracked, thereby reducing the error accumulation caused by the angle deviation, and in the case of weak or no GPS signal, it can still maintain a high trajectory restoration accuracy.

[0106] In order to facilitate those skilled in the art to understand the present solution, the specific implementation process of the method for determining the estimated position point set described in steps S102-S105 provided in the embodiment of the present application is described more clearly and completely below: Figure 5 As shown in the figure, the road network data includes road sections L1, L2 and L3, and G1 is a GPS track point. Figure 5 The GPS track point set in includes GPS track points G1 to G4 ( Figure 5 Taking G2 to G4 as an example, for GPS track point G1, executing steps S102-S105 to obtain multiple estimated position points corresponding to G1 includes:

[0107] First, based on the above step S102, for G1, sections L1 and L2 are determined as candidate sections of G1, and G1 is projected onto each candidate section to obtain reference points P1-1 and P1-2.

[0108] Then, based on the above step S103, for the reference point P1-1, taking the reference point P1-1 as the starting point, combined with the direction of G1 and the speed from G1 to the next trajectory point G2 (G2 is not shown in the figure), the estimated position point S2-1 is obtained. Similarly, for the reference point P1-2, the estimated position point S2-2 is obtained.

[0109] Next, based on the above step S104, for the reference point P1-1, taking the reference point P1-1 as the starting point, combined with the driving direction of the road section L1, and the speed and time difference from G1 to the next trajectory point G2, a new reference point P2-1 is obtained. Similarly, for the reference point P1-2, a new reference point P2-2 is obtained.

[0110] Finally, based on the above step S105, the above steps S103 and S104 are repeatedly executed, including: based on the reference points P2-1 and P2-2, the estimated position points S3-1 and S3-2, and new reference points P3-1 and P3-2 (P3-2 is not shown in the figure); based on the reference points P3-1 and P3-2, the estimated position points S4-1 and S4-2 (P4-2 is not shown in the figure), and new reference points P4-1 and P4-2 (P4-2 is not shown in the figure) are obtained. If there is no next GPS track point G5, the loop is stopped. All track points obtained during the loop are integrated to obtain multiple estimated position points corresponding to the track point G1.

[0111] The above steps S102-S105 are executed for each GPS track point to obtain a set of estimated position points.

[0112] In the embodiment of the present application, the method for determining the estimated position point proposed in steps S102-S105, by combining the vehicle's driving direction and speed information, calculates the estimated position point from the known GPS trajectory point to the driving direction. This method can effectively reduce the cumulative error caused by angle deviation, especially when the GPS signal is weak or there is no signal, and can still maintain a high trajectory restoration accuracy.

[0113] For example, refer to Figure 6 As shown in the figure, the road sections L1 to L10 are included, GPS track points G1 and G5 are correct GPS positions, while GPS track points G2, G3, and G4 are abnormal GPS positions, indicating that the vehicle may have passed through some complex environments (such as tunnels, under viaducts, urban areas with tall buildings, etc.) at that time. At this time, based on GPS track point G1, the estimated position points S2-1, S3-1, S4-1, and S5-1 can still be obtained according to the estimated position point determination method proposed by the method, so that the method can still obtain the trajectory restoration result of L1+L2+L3+L4 later.

[0114] In the above step S106, based on the road network data, the matching probability and evolution probability associated with each to-be-matched trajectory point in the GPS trajectory point set, the dead reckoning point set and the estimated position point set are calculated to obtain the trajectory restoration result. Specifically, the following steps S1061-S1066 are included:

[0115] S1061: Based on the position of each road segment in the road network data, determine a plurality of candidate road segments corresponding to each trajectory point to be matched.

[0116] The track points to be matched represent all points in the GPS track point set, the dead reckoning point set, and the estimated position point set.

[0117] Specifically, the corresponding second preset proximity range may be determined based on the trajectory point to be matched, and if a road section in the road network data is within the second preset proximity range, the road section is determined as a candidate road section corresponding to the trajectory point to be matched.

[0118] The second preset proximity range can be determined in the following two ways: the first way is to set the second preset radius with the GPS track point as the center of the circle, and the range of the circle is the second preset proximity range. The second way is to set the side length of the second preset rectangle with the GPS track point as the center, and determine the rectangular area along the longitude and latitude lines according to the side length of the second preset rectangle, and the rectangular area is the second preset proximity range. At the same time, in some cases, the first preset radius can be equal to the second preset radius, and the side length of the first preset rectangle can be equal to the side length of the second preset rectangle.

[0119] Those skilled in the art may also select a method for determining the second preset proximity range according to actual conditions, and no specific limitation is made in the embodiments of the present application.

[0120] S1062: Combine each to-be-matched trajectory point and each corresponding candidate road segment into a candidate matching pair.

[0121] In order to facilitate those skilled in the art to understand the present solution, the candidate matching pairs provided in the embodiment of the present application are described more clearly and completely below: Figure 7 As shown, the road network data includes L1, L2 and L3, G1 to G3 are GPS track points, D1 to D3 are dead reckoning points, S2-1 and S2-2 are dead reckoning points. Please note that this is to clearly demonstrate the implementation effect of this method. Figure 7 Only some of the estimated positions are shown.

[0122] by Figure 7 For example, each trajectory point to be matched in the figure and each corresponding candidate road section can form a candidate matching pair "G1->L1", "G1->L2", "D1->L1", "G2->L2", "G2->L3", "D2->L3", "S2-1->L2", "S2-1->L3", "S2-2->L2", "S2-2->L3", "G3->L3", "D2->L3", "D3->L3".

[0123] S1063: Calculate the matching probability and multiple associated evolution probabilities corresponding to each candidate matching pair according to the distance between the to-be-matched trajectory point and the candidate road segment in the candidate matching pair. Specifically, the following steps S10631-S10633 are included:

[0124] S10631: Calculate the matching probability of each candidate matching pair according to the projection distance between the to-be-matched trajectory point and the candidate road segment in the candidate matching pair.

[0125] Specifically, the matching probability of each candidate matching pair may be calculated based on the following formula 3 according to the projection distance between the to-be-matched trajectory point and the candidate road segment in the candidate matching pair:

[0126]

[0127] Where P1 represents the matching probability, ranging from 0 to 1, e is a natural constant, distance is the projection distance, and σ is the standard deviation, which determines the amplitude of the distribution. Researchers are required to set empirical values ​​according to the distribution of the trajectory points to be matched.

[0128] At the same time, the matching probability needs to follow the Gaussian distribution (normal distribution), such as Figure 8 As shown, the y-axis is the matching probability and the x-axis is the projection distance. It can be seen that the smaller the projection distance, the closer the matching probability is to 1, and the larger the projection distance, the closer the matching probability is to 0.

[0129] For example, see Fig. 9 As shown, G1 is the GPS trajectory point, P1-1 is the projection point of G1 to L1, and P1-2 is the projection point of G1 to L2. Since the distance from G1 to P1-1 is smaller than the distance from G1 to P1-2, the matching probability of the candidate matching pair "G1->L1" will be greater than the matching probability of the candidate matching pair "G1->L2".

[0130] S10632: According to the timestamps of the to-be-matched trajectory points in each candidate matching pair, all candidate matching pairs are divided into a plurality of time point candidate sets sorted in order of time from earliest to latest.

[0131] In order to facilitate those skilled in the art to understand the present solution, the following is a clearer and more complete description of the candidate time point set provided in the embodiment of the present application: Figure 7 As shown, according to the above step S1062, candidate matching pairs "G1->L1", "G1->L2", "D1->L1", "G2->L2", "G2->L3", "D2->L3", "S2-1->L2", "S2-1->L3", "S2-2->L2", "S2-2->L3", "G3->L3", "D2->L3", and "D3->L3" can be formed.

[0132] Among them, G1 and D1 belong to the first timestamp, G2, D2, S2-1 and S2-2 belong to the second timestamp, G3 and D3 belong to the third timestamp, and the three timestamps are sorted by time.

[0133] Therefore, all candidate matching pairs can be divided into three time point candidate sets according to the timestamp:

[0134] The candidate set at time point 1 includes "G1->L1", "G1->L2", and "D1->L1".

[0135] The candidate set for time point 2 includes "G2->L2", "G2->L3", "D2->L3", "S2-1->L2", "S2-1->L3", "S2-2->L2", and "S2-2->L3".

[0136] The candidate set at time point 1 includes "G3->L3", "D2->L3", and "D3->L3".

[0137] S10633: For each candidate matching pair in the candidate set at each time point, calculate the evolution probability between the candidate matching pair and all candidate matching pairs in the candidate set at the next time point.

[0138] Specifically, for each candidate matching pair in the candidate set at each time point, the evolution probability between the candidate matching pair and all candidate matching pairs in the candidate set at the next time point is calculated. For ease of understanding, any candidate matching pair in the candidate set at any time point is taken as the first candidate matching pair, and any candidate matching pair in the candidate set at the next time point of the first candidate matching pair is taken as the second candidate matching pair.

[0139] The method for calculating the evolution probability between the first candidate matching pair and the second candidate matching pair specifically includes: first, calculating the distance between two to-be-matched trajectory points in the first candidate matching pair and the second candidate matching pair, recorded as the first distance, then calculating the distance between the to-be-matched trajectory points in the first candidate matching pair and the second candidate matching pair to two projection points of the candidate road section, respectively, recorded as the second distance, and calculating the difference between the first distance and the second distance, recorded as the distance difference, and finally, calculating the evolution probability between the first candidate matching pair and the second candidate matching pair based on the following formula 4:

[0140]

[0141] Where P2 represents the evolution probability, ranging from 0 to 1, e is a natural constant, distance difference is the distance difference, and σ is the standard deviation, which determines the amplitude of the distribution. Researchers are required to set empirical values ​​according to the distribution of the trajectory points to be matched, which will not be elaborated here.

[0142] For example, see Fig.10As shown, G1 and G2 are GPS trajectory points, P1-1 is the projection point of G1 to L1, P1-2 is the projection point of G1 to L2, P2-1 is the projection point of G2 to L3, and P2-2 is the projection point of G2 to L2. Since the distance from P1-1 to P2-1 is closer to the distance from G1 to G2 than P1-1 to P2-2, P1-2 to P2-1, and P1-2 to P2-2, the evolution probability between the candidate matching pair "G1->P1-1" and the candidate matching pair "G2->P2-1" is the largest.

[0143] In order to facilitate those skilled in the art to understand the present solution, the following is a more complete description of the candidate time point set provided in the embodiment of the present application: Figure 7 In the example of the road network shown in FIG. 1 , three candidate time point sets are obtained. After completing the calculation steps of the matching probability and the evolution probability in the above steps S10631 and S10633, a probability diagram between all candidate matching pairs can be constructed, as shown in FIG. Fig.11 As shown, there are three time point candidate sets, including 3, 7, and 3 candidate matching pairs respectively. The probability value in the brackets after the candidate matching pairs is the matching probability, and the probability value on the arrow between two candidate matching pairs is the evolution probability.

[0144] In the embodiment of the present application, after executing the above step S1063, the following step is further included:

[0145] S107: Calculate the probability of each candidate matching pair according to the matching probability corresponding to each candidate matching pair and the associated multiple evolution probabilities, and after calculating the probabilities of all candidate matching pairs, select and retain multiple candidate matching pairs with the highest probability. Specifically, the following steps S1071-S1074 are included:

[0146] S1071: For each candidate matching pair in the candidate set at each time point, connect the candidate matching pair with each candidate matching pair in the candidate set at the previous time point to obtain multiple ternary arrangements corresponding to the candidate matching pair.

[0147] S1072: Calculate the probability of each ternary arrangement according to the matching probability of each candidate matching pair in the ternary arrangement and the evolution probability between adjacent candidate matching pairs.

[0148] Specifically, the probability of the ternary arrangement may be obtained by multiplying the matching probability of two candidate matching pairs in each ternary arrangement and the evolution probability between the two candidate matching pairs.

[0149] S1073: Select the maximum value of the probabilities among all the ternary arrangements corresponding to the candidate matching pair as the probability of the candidate matching pair.

[0150] S1074: According to the probabilities of all candidate matching pairs in the candidate set at each time point, retain a preset threshold number of candidate matching pairs with the highest probability in the candidate set at each time point.

[0151] In order to facilitate those skilled in the art to understand the present solution, the specific implementation process of the above steps S1071-S1074 provided in the embodiment of the present application is described more clearly and completely below: Fig.12 As shown, there are 2 time point candidate sets, each including 2 candidate matching pairs. The probability value in brackets after the candidate matching pairs is the matching probability, the probability value on the arrow between the two candidate matching pairs is the evolution probability, and the ternary arrangements include "G1->L1"->"G2->L2", "G1->L1"->"G2->L3", "G1->L2"->"G2->L2" and "G1->L2"->"G2->L3".

[0152] For the candidate matching pair “G2->L2”, the corresponding two triple permutations are “G1->L1”->“G2->L2” and “G1->L2”->“G2->L2”.

[0153] The probability of the ternary arrangement "G1->L1"->"G2->L2" = 90% (matching probability of G1->L1) × 91% (evolution probability between "G1->L1" and "G2->L2") × 95% (matching probability of G2->L2) = 77.805%.

[0154] Similarly, it can be calculated that the probability of the ternary arrangement "G1->L2"->"G2->L2" is equal to 68.4%.

[0155] The probability of the candidate matching pair "G2->L2" = MAX(77.805%, 68.4%) = 77.805%.

[0156] Likewise, it can be calculated that the probability of the candidate matching pair "G2->L3" is equal to 79.38%.

[0157] When the preset candidate threshold number is 1, only the candidate matching pair "G2->L3" will be retained, and all data associated with the candidate matching pair "G2->L2" will be deleted.

[0158] In the embodiment of the present application, the operations performed in the above steps S1071-S1074 can effectively optimize resource utilization. By limiting the maximum number of candidates in the candidate set at each time point and retaining multiple previous candidate matching pairs with the highest probability, effective utilization of computing resources is ensured while ensuring high-quality output of trajectory restoration.

[0159] S1064: Construct multiple candidate point sequences according to the timestamps of the to-be-matched trajectory points in each candidate matching pair.

[0160] Specifically, any candidate matching pair may be selected from each time point candidate set, a candidate point sequence may be constructed in order, and the operation of constructing the candidate point sequence may be repeated until all time point candidate sets are traversed to obtain all possible candidate point sequences, thereby obtaining multiple candidate point sequences.

[0161] Taking the above example Fig.11 Taking the three time point candidate sets in as an example, the three time point candidate sets include 3, 7, and 3 candidate matching pairs respectively. By executing the above step S1064, a total of 91 candidate point sequences can be obtained, such as "G1->L1"->"G2->L2"->"G3->L3", "D1->L1"->"D2->L3"->"D3->L3".

[0162] In the embodiment of the present application, after executing the above step S1064, it is also necessary to execute step S108 to delete some unreasonable data. S108 specifically includes the following steps S1081-S1082:

[0163] S1081: In a candidate point sequence, if the distance between the to-be-matched trajectory points in two adjacent candidate matching pairs is greater than the product of the time difference between the two to-be-matched trajectory points and a preset maximum driving speed, the candidate point sequence is deleted.

[0164] Specifically, in a candidate point sequence, if the distance between the to-be-matched trajectory points in two adjacent candidate matching pairs is greater than the product of the time difference between the two to-be-matched trajectory points and the preset maximum driving speed, it is considered unreachable and the candidate point sequence is deleted. The preset maximum driving speed can be set to 200 km / h.

[0165] S1082: In a candidate point sequence, if there is no topological relationship between the candidate road segments in two adjacent candidate matching pairs, the candidate point sequence is deleted.

[0166] For example, see Fig.13 As shown in the figure, the road sections L1-L2 and L3-L4 are not connected and have no topological relationship. The candidate road section of the GPS track point G1 is only L1, and the candidate road sections of G2 are L2 and L3. However, since L1 and L3 have no topological relationship, it is necessary to delete the candidate point sequence "G1->L1"->"G2->L3".

[0167] In the embodiment of the present application, the operation of deleting unreasonable data performed in the above steps S1081-S1082 can delete unreasonable or low-probability candidate points, avoid the exponential growth of the number of candidate sets, significantly improve computing efficiency, and shorten the time required for trajectory restoration.

[0168] S1065: Calculate the probability of each candidate point sequence according to the matching probability corresponding to each candidate matching pair and the associated multiple evolution probabilities.

[0169] Specifically, the probability of each candidate point sequence may be obtained by multiplying the matching probability of each candidate matching pair in the candidate point sequence and the evolution probability between each two adjacent candidate matching pairs.

[0170] For example, based on the example in S1063 above Fig.11 As shown, the probability of the candidate point sequence "G1->L1"->"G2->L2"->"G3->L3" = 90% (matching probability of G1->L1) × 91% (evolution probability between "G1->L1" and "G2->L2") × 95% (matching probability of G2->L2) × 0% (evolution probability between "G2->L2" and "G3->L3") × 92% (matching probability of G3->L3) = 0%.

[0171] The probability of the candidate point sequence "D1->L1"->"D2->L3"->"D3->L3" = 99% (matching probability of D1->L1) × 99% (evolution probability between "D1->L1" and "D2->L3") × 99% (matching probability of D2->L3) × 99% (evolution probability between "D2->L3" and "D3->L3") × 99% (matching probability of D3->L3) = 95%.

[0172] S1066: Select the candidate point sequence with the highest probability to obtain the trajectory restoration result.

[0173] by Fig.11 Taking the candidate matching pair probability diagram shown in the figure as an example, it can be calculated that the candidate point sequence with the highest probability is "D1->L1"->"D2->L3"->"D3->L3", thus, it can be determined that the trajectory restoration result is L1+L2+L3.

[0174] Embodiment 2

[0175] Based on the same inventive concept, the embodiment of the present invention also provides a vehicle trajectory restoration device based on multi-source data, referring to Fig.14 As shown, the device comprises:

[0176] The first acquisition module 101 is used to acquire a GPS track point set, a dead reckoning point set and a road network data; the road network data includes a plurality of road sections; the GPS track point set includes a plurality of GPS track points sorted in time from earliest to latest;

[0177] A reference point determination module 102 is used to determine multiple reference points on multiple road sections for each GPS track point;

[0178] The estimated position module 103 is used to calculate a estimated position point for each reference point, taking the reference point as a starting point and combining the direction and speed from the GPS track point to the next GPS track point;

[0179] The reference point update module 104 is used to calculate a new reference point based on the reference point as the starting point, the driving direction of the road section where the reference point is located, the time difference from the GPS track point to the next GPS track point, and the speed;

[0180] A first combining module 105 is used to take the new reference point as the starting point, repeat the steps of calculating the estimated position point and updating the reference point by the estimated position module and the updating reference point module, until there is no next GPS track point, and combine all the estimated position points to obtain an estimated position point set;

[0181] The first calculation module 106 is used to calculate the matching probability and the evolution probability associated with each to-be-matched trajectory point in the GPS trajectory point set, the dead reckoning point set and the estimated position point set based on the road network data to obtain a trajectory restoration result.

[0182] Embodiment 3

[0183] Based on the same inventive concept, an embodiment of the present invention further provides a computer-readable storage medium on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the driving trajectory restoration method based on multi-source data as described in the above embodiment 1 is implemented.

[0184] Embodiment 4

[0185] Based on the same inventive concept, an embodiment of the present invention further provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the method for restoring a vehicle trajectory based on multi-source data as described in the first embodiment above.

[0186] Embodiment 5

[0187] Based on the same inventive concept, an embodiment of the present invention further provides a computer device, including a memory, a processor, and a computer program stored in the memory. When the processor executes the computer program, the driving trajectory restoration method based on multi-source data as described in the above embodiment 1 is implemented.

[0188] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0189] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0190] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0191] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0192] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A vehicle trajectory restoration method based on multi-source data, characterized in that: include: Acquire a GPS track point set, a dead reckoning point set and a road network data; the road network data includes a plurality of road sections; the GPS track point set includes a plurality of GPS track points sorted in time from earliest to latest; For each GPS track point, multiple reference points are determined on multiple road sections; For each reference point, taking the reference point as the starting point, a calculated position point is calculated by combining the direction and speed from the GPS track point to the next GPS track point; Taking the reference point as the starting point, a new reference point is calculated based on the driving direction of the road section where the reference point is located, the time difference from the GPS track point to the next GPS track point, and the speed; Taking the new reference point as the starting point, repeat the above steps of calculating the estimated position point and updating the reference point until there is no next GPS track point, and combine all the estimated position points to obtain an estimated position point set; Based on the road network data, the matching probability and the evolution probability associated with each to-be-matched trajectory point in the GPS trajectory point set, the dead reckoning point set and the estimated position point set are calculated to obtain a trajectory restoration result.

2. The method according to claim 1, characterized in that For each GPS track point, multiple reference points are determined on multiple road sections, including: For each GPS track point, based on the position of each road segment in the road network data, determine a plurality of candidate road segments corresponding to the GPS track point; The GPS track points are projected onto each candidate road section to obtain multiple reference points.

3. The method according to claim 1, characterized in that For each reference point, taking the reference point as the starting point, and combining the direction and speed from the GPS track point to the next GPS track point, a calculated position point is calculated, including: Calculate the estimated distance based on the time difference and speed from the GPS track point to the next GPS track point; For each reference point, the reference point is used as a starting point, and the latitude and longitude coordinates of the estimated position point are calculated according to the latitude and longitude coordinates of the reference point, the direction and the estimated distance.

4. The method according to claim 1, characterized in that: The method of calculating the matching probability and the evolution probability associated with each to-be-matched trajectory point in the GPS trajectory point set, the dead reckoning point set, and the estimated position point set based on the road network data to obtain a trajectory restoration result includes: Based on the position of each road section in the road network data, determining a plurality of candidate road sections corresponding to each to-be-matched trajectory point; Combining each to-be-matched trajectory point with each corresponding candidate road segment into a candidate matching pair; Calculating the matching probability and associated multiple evolution probabilities corresponding to each candidate matching pair according to the distance between the to-be-matched trajectory point and the candidate road section in the candidate matching pair; Constructing multiple candidate point sequences according to the timestamps of the to-be-matched trajectory points in each candidate matching pair; Calculate the probability of each candidate point sequence according to the matching probability corresponding to each candidate matching pair and the associated multiple evolution probabilities; The candidate point sequence with the highest probability is selected to obtain the trajectory restoration result.

5. The method according to claim 4, characterized in that After obtaining multiple candidate point sequences, it also includes: In a candidate point sequence, if the distance between the to-be-matched trajectory points in two adjacent candidate matching pairs is greater than the product of the time difference between the two to-be-matched trajectory points and the preset maximum driving speed, then the candidate point sequence is deleted; and / or In a candidate point sequence, if there is no topological relationship between the candidate road segments in two adjacent candidate matching pairs, the candidate point sequence is deleted.

6. The method according to claim 4, characterized in that The step of calculating the matching probability and the associated multiple evolution probabilities corresponding to each candidate matching pair according to the distance between the to-be-matched trajectory point and the candidate road segment in the candidate matching pair includes: Calculating the matching probability of each candidate matching pair according to the projection distance between the to-be-matched trajectory point and the candidate road section in the candidate matching pair; According to the timestamps of the to-be-matched trajectory points in each candidate matching pair, all candidate matching pairs are divided into a plurality of time point candidate sets sorted in order of time from earliest to latest; For each candidate matching pair in the candidate set at each time point, the evolution probability between the candidate matching pair and all candidate matching pairs in the candidate set at the next time point is calculated.

7. The method according to claim 6, characterized in that According to the timestamps of the to-be-matched trajectory points in each candidate matching pair, multiple candidate point sequences are constructed, including: Take any candidate matching pair from each time point candidate set, construct a candidate point sequence in order, and repeat the operation of constructing the candidate point sequence until all time point candidate sets are traversed to obtain all possible candidate point sequences, thereby obtaining the multiple candidate point sequences.

8. The method according to claim 7, characterized in that Calculating the probability of each candidate point sequence according to the matching probability corresponding to each candidate matching pair and the associated multiple evolution probabilities, including: The probability of each candidate point sequence is calculated according to the matching probability of each candidate matching pair in the candidate point sequence and the evolution probability between each two adjacent candidate matching pairs.

9. The method according to claim 8, characterized in that After all candidate matching pairs are divided into multiple time point candidate sets sorted from earliest to latest, it also includes: For each candidate matching pair in the candidate set at each time point, connect the candidate matching pair with each candidate matching pair in the candidate set at the previous time point to obtain multiple ternary permutations corresponding to the candidate matching pair; Calculating the probability of each ternary arrangement according to the matching probability of each candidate matching pair in the ternary arrangement and the evolution probability between adjacent candidate matching pairs; Selecting the maximum value of the probabilities among all the ternary arrangements corresponding to the candidate matching pair as the probability of the candidate matching pair; According to the probabilities of all candidate matching pairs in the candidate set at each time point, a preset candidate threshold number of candidate matching pairs with the highest probabilities in the candidate set at the time point is retained.

10. A vehicle trajectory restoration device based on multi-source data, characterized in that: include: A first acquisition module is used to acquire a GPS track point set, a dead reckoning point set and a road network data; the road network data includes a plurality of road sections; the GPS track point set includes a plurality of GPS track points sorted in time from earliest to latest; A reference point determination module is used to determine multiple reference points on multiple road sections for each GPS track point; The estimated position module is used to calculate a estimated position point for each reference point, taking the reference point as a starting point and combining the direction and speed from the GPS track point to the next GPS track point; The reference point update module is used to calculate a new reference point based on the reference point as the starting point, the driving direction of the road section where the reference point is located, the time difference from the GPS track point to the next GPS track point, and the speed; A first combining module is used to take the new reference point as a starting point, repeat the steps of calculating the estimated position point and updating the reference point by the estimated position module and the updating reference point module, until there is no next GPS track point, and combine all the estimated position points to obtain an estimated position point set; The first calculation module is used to calculate the matching probability and evolution probability associated with each to-be-matched trajectory point in the GPS trajectory point set, the dead reckoning point set and the estimated position point set based on the road network data to obtain a trajectory restoration result.

11. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instruction is executed by a processor, the driving trajectory restoration method based on multi-source data described in any one of claims 1 to 9 is implemented.

12. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the driving trajectory restoration method based on multi-source data described in any one of claims 1 to 9 is implemented.

13. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the driving trajectory restoration method based on multi-source data as described in any one of claims 1-9.

Images