A trajectory matching method combining target motion feature joint discrimination

Abstract

The application discloses a trajectory matching method combining target motion characteristics and joint discrimination, and relates to the field of geographic information processing. The application randomly generates trajectory starting points in three sub-regions, generates a simulation trajectory set according to the trajectory starting points and simulation trajectory parameters, obtains a positive and negative sample set in a cross matching mode by introducing noise, resamples the positive and negative sample set to obtain a simulation trajectory training matching sample pair set, calculates distance features and direction features of the simulation trajectory training matching sample pair, inputs the distance features and the direction features into an SVM classifier, constructs a trajectory matching classification model, inputs two trajectories to be matched into the trajectory matching classification model, and obtains a corresponding matching result. The application can effectively utilize target motion element information in the trajectory, construct a matching model, realize matching of a target trajectory, and is more accurate and reliable in the trajectory matching result, realizes matching of a most similar path target trajectory, and thus provides strong technical support for optimization and management of marine transportation.

Description

Technical Field

[0001] This invention relates to the field of geographic information processing, and in particular to a trajectory matching method that combines target motion characteristics for joint discrimination. Background Technology

[0002] Maritime transport, as a vital link connecting the world, has an increasingly significant impact on economic activities. Ships, as the primary means of maritime transport, require careful management and optimization of their operational trajectories to ensure logistical efficiency and safety. Ship target trajectory matching technology plays a crucial role not only in shipping logistics, port management, and maritime safety, but also has a profound impact on marine resource development and environmental monitoring. However, with the increasing number of ships and the increasing complexity of shipping routes, traditional trajectory matching methods face significant challenges. Therefore, developing an efficient ship target trajectory matching method is of great socio-economic significance for improving the overall efficiency and safety of maritime transport.

[0003] Existing ship trajectory matching algorithms primarily rely on simple geometric relationships or time series analysis, often neglecting the complexity and diversity of ship motion characteristics. When processing large-scale, highly dynamic trajectory data, these algorithms frequently exhibit low accuracy and weak robustness. Specifically, existing algorithms often fail to accurately capture key motion characteristics such as ship speed, acceleration, and steering during the matching process, limiting the reliability and practicality of the matching results. Therefore, there is an urgent need for methods that utilize target motion feature characteristics for target trajectory matching. Summary of the Invention

[0004] In view of this, the present invention proposes a trajectory matching method that combines target motion characteristics for joint discrimination. This method comprehensively considers the target's motion elements such as velocity, acceleration, and direction, thereby improving the reliability of target trajectory matching.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A trajectory matching method that combines target motion features for joint discrimination includes the following steps:

[0007] Step 1: Set up three sub-regions of different scales for trajectory data of different densities within the simulation region R, denoted as R = {r1, r2, r3};

[0008] Step 2: Randomly generate m trajectory starting points in each sub-region of the simulation region R, and construct a trajectory starting point set. i = 1, 2, 3; m ≥ 7;

[0009] Step 3, based on the set of trajectory starting points By combining trajectory simulation parameters, a set of simulation trajectories is generated.

[0010] Step 4, for the set of simulation trajectories For each simulated trajectory, random noise at two scales is introduced to obtain two simulated trajectory matching samples at different scales. These are recorded as a simulated trajectory matching sample pair, and finally, a set of simulated trajectory matching sample pairs is generated.

[0011] Step 5: Set up the simulation trajectory matching sample pairs. All simulated trajectory matching sample pairs are labeled as positive sample pairs and given a label of 1, thus obtaining the set of positive sample pairs.

[0012] Step 6: Set up the simulation trajectory matching sample pairs Cross-matching any two different simulated trajectory matching sample pairs yields multiple results that do not belong to the set of simulated trajectory matching sample pairs. From the simulated trajectory matching sample pairs, arbitrarily select 3m simulated trajectory matching sample pairs, label them as negative sample pairs, and assign them a label of 0, thus obtaining the set of negative sample pairs.

[0013] Step 7, for the set of positive sample pairs and negative sample pairs set Resampling is performed to construct a set of training matching sample pairs for the simulated trajectory.

[0014] Step 8: Calculate the set of training matching sample pairs for the simulation trajectory. In each simulated trajectory, distance features are trained on matching sample pairs to obtain a distance feature set.

[0015] Step 9: Calculate the set of training matching sample pairs for the simulation trajectory. The orientation features of each simulated trajectory are trained and matched with sample pairs to obtain the orientation feature set.

[0016] Step 10: Set up training matching sample pairs for the simulation trajectory. Distance feature set and directional feature set The input is fed into an SVM classifier to build a trajectory matching classification model M. SVM The trajectory matching classification model M SVM The input is the distance and direction features between two trajectories, and the output is a label 0 or 1;

[0017] Step 11, for the two trajectories F to be matched re and S reThe distance and direction features between the two are calculated, and then the distance and direction features are input into the trajectory matching classification model M. SVM In the middle, the trajectory F is obtained. re and S re The corresponding label, if the label is 0, indicates trajectory F. re and S re A mismatch occurs if the label is 1, indicating trajectory F. re and S re Matching is achieved by combining the target's motion features for trajectory matching.

[0018] Furthermore, the set of simulation trajectories in step 3 It can be represented as:

[0019]

[0020] in, Represents the set of simulation trajectories The j-th simulated trajectory in the dataset includes three trajectory parameters and a sampling time interval. and These represent the time intervals corresponding to the three trajectory segments respectively. For different simulation trajectories, the sum of the time intervals corresponding to the three trajectory segments is the same. and These represent the initial velocities for the corresponding time intervals. and These represent the initial directions for the corresponding time intervals. and These represent the unit-time velocity variables for the corresponding time intervals. and These represent the unit time direction variables for the corresponding time intervals; Δt represents the sampling time interval of the trajectory points, which is the same for different simulation trajectories.

[0021] Furthermore, in step 4, the set of simulation trajectory matching sample pairs Specifically:

[0022]

[0023] in, For the set of simulation trajectory matching sample pairs The j-th simulated trajectory matching sample pair: includes two simulated trajectory matching samples at different scales. and They are all based on simulated trajectories Adding positional noise generates the following: and They represent the simulation trajectory respectively The generated noise mean at different scales, and They represent the simulation trajectory respectively The standard deviations generated at different scales.

[0024] Furthermore, in step 5, the set of positive sample pairs... Specifically:

[0025]

[0026] Furthermore, in step 6, the negative sample pair set Specifically:

[0027]

[0028] in, negative sample pair set The k-th negative sample pair: including and Two simulated trajectory matching samples.

[0029] Furthermore, step 7 includes the following steps:

[0030] Step 701, set the positive sample pairs and negative sample pairs set The elements are merged, and in the new merged set, each element belongs to... or

[0031] Step 702: Each element obtains a corresponding simulated trajectory training matching sample pair based on its own contained trajectory information.

[0032]

[0033] in, and They represent Trajectory information of two simulated trajectory matching samples in the corresponding elements; express The tag of the corresponding element, if The corresponding element belongs to but like The corresponding element belongs to but and They represent The longitude, latitude, and instantaneous heading of the first trajectory point corresponding to the simulated trajectory matching sample. and They represent The longitude, latitude, and instantaneous heading of the first trajectory point in the corresponding simulation trajectory matching sample, where n represents... and The number of trajectory points corresponding to the simulated trajectory matching samples, the value of n is determined by the set of simulated trajectories. The sum of the time intervals corresponding to the three trajectories and the sampling time interval Δt are determined;

[0034] Step 703 finally yields the set of training matching sample pairs for the simulated trajectory.

[0035]

[0036] Furthermore, in step 8, the distance feature set of the simulated trajectory training matching sample pairs is... Specifically:

[0037]

[0038] in, Representing the distance feature set The distance features of the training matching sample pair of the l-th simulated trajectory are calculated. hist() represents solving the histogram statistical frequency, and log() represents calculating the logarithm.

[0039] Furthermore, in step 9, the simulated trajectory training matches the direction feature set of the sample pairs. Specifically:

[0040]

[0041] in, Representation of directional feature set The directional features of the training matching sample pair of the l-th simulated trajectory.

[0042] Furthermore, the trajectory F to be matched in step 11 re and S re Specifically:

[0043] F re ={x F (1),y F (1),w F (1)),(x F (2),y F (2),w F (2)),...,

[0044] (x F (n),y F (n),w F (n))}

[0045] Sre ={x S (1),y S (1),w S (1)),(x S (2),y S (2),w S (2)),...,

[0046] (x S (n),y S (n),w S (n))}

[0047] Where, x F (1) y F (1) and w F (1) Represent the trajectory F respectively re The longitude, latitude, and instantaneous heading of the first trajectory point, x S (1) y S (1) and w S (1) Represent the simulation trajectory S respectively re The longitude, latitude, and instantaneous heading of the first trajectory point, and the simulated trajectory F. re and S re The number of trajectory points is n.

[0048] Due to the adoption of the above technical solution, the beneficial effects of this invention compared with the prior art are as follows:

[0049] 1. This invention proposes a trajectory matching method that combines target motion features for joint discrimination. It can effectively utilize the target motion element information in the trajectory to construct a matching model and achieve target trajectory matching.

[0050] 2. This invention can provide more accurate and reliable trajectory matching results, achieving the matching of the most similar path target trajectory, thereby providing strong technical support for the optimization and management of marine transportation. Attached Figure Description

[0051] Figure 1 This is an overall flowchart of a trajectory matching method that combines target motion features for joint discrimination in an embodiment of the present invention. Detailed Implementation

[0052] The invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0053] A trajectory matching method that combines target motion features for joint discrimination, such as... Figure 1 As shown, it includes the following steps;

[0054] Step 1: Set up three sub-regions of different scales for trajectory data of different densities within the simulation region R, denoted as R = {r1, r2, r3};

[0055] Specifically, subregion r i It can be represented as:

[0056]

[0057] In the formula, Representing regions r respectively i The starting point's longitude and latitude, Representing regions r respectively i The longitude and latitude variables.

[0058] Starting point subregion r i It is a rectangular region; therefore, the coordinates of the other diagonal endpoint of this region are as follows:

[0059]

[0060] Step 2: Randomly generate m trajectory starting points in each sub-region of the simulation region R, and construct a trajectory starting point set. i = 1, 2, 3; m ≥ 7;

[0061] Specifically, Set of starting points of the trajectory The j-th starting point in the region, and it falls within the subregion r. i Inside.

[0062] Step 3, based on the set of trajectory starting points By combining trajectory simulation parameters, a set of simulation trajectories is generated.

[0063] Step 4, for the set of simulation trajectories For each simulated trajectory, random noise at two scales is introduced to obtain two simulated trajectory matching samples at different scales. These are recorded as a simulated trajectory matching sample pair, and finally, a set of simulated trajectory matching sample pairs is generated.

[0064] Step 5: Set up the simulation trajectory matching sample pairs. All simulated trajectory matching sample pairs are labeled as positive sample pairs and given a label of 1, thus obtaining the set of positive sample pairs.

[0065] Step 6: Set up the simulation trajectory matching sample pairs Cross-matching any two different simulated trajectory matching sample pairs yields multiple results that do not belong to the set of simulated trajectory matching sample pairs. From the simulated trajectory matching sample pairs, arbitrarily select 3m simulated trajectory matching sample pairs, label them as negative sample pairs, and assign them a label of 0, thus obtaining the set of negative sample pairs.

[0066] Step 7, for the set of positive sample pairs and negative sample pairs set Resampling is performed to construct a set of training matching sample pairs for the simulated trajectory.

[0067] Step 8: Calculate the set of training matching sample pairs for the simulation trajectory. In each simulated trajectory, distance features are trained on matching sample pairs to obtain a distance feature set.

[0068] Step 9: Calculate the set of training matching sample pairs for the simulation trajectory. The orientation features of each simulated trajectory are trained and matched with sample pairs to obtain the orientation feature set.

[0069] Step 10: Set up training matching sample pairs for the simulation trajectory. Distance feature set and directional feature set The input is fed into an SVM classifier to build a trajectory matching classification model M. SVM The trajectory matching classification model M SVM The input is the distance and direction features between two trajectories, and the output is a label 0 or 1;

[0070] Step 11, for the two trajectories F to be matched re and S re The distance and direction features between the two are calculated, and then the distance and direction features are input into the trajectory matching classification model M. SVM In the middle, the trajectory F is obtained. re and S re The corresponding label, if the label is 0, indicates trajectory F. re and S re A mismatch occurs if the label is 1, indicating trajectory F. re and S re Matching is achieved by combining the target's motion features for trajectory matching.

[0071] Furthermore, the set of simulation trajectories in step 3 It can be represented as:

[0072]

[0073] in, Represents the set of simulation trajectories The j-th simulated trajectory in the dataset includes three trajectory parameters and a sampling time interval. and These represent the time intervals corresponding to the three trajectory segments respectively. For different simulation trajectories, the sum of the time intervals corresponding to the three trajectory segments is the same. and These represent the initial velocities for the corresponding time intervals. and These represent the initial directions for the corresponding time intervals. and These represent the unit-time velocity variables for the corresponding time intervals. and These represent the unit time direction variables for the corresponding time intervals; Δt represents the sampling time interval of the trajectory points, which is the same for different simulation trajectories.

[0074] Furthermore, in step 4, the set of simulation trajectory matching sample pairs Specifically:

[0075]

[0076] in, For the set of simulation trajectory matching sample pairs The j-th simulated trajectory matching sample pair: includes two simulated trajectory matching samples at different scales. and They are all based on simulated trajectories Adding positional noise generates the following: and They represent the simulation trajectory respectively The generated noise mean at different scales, and They represent the simulation trajectory respectively The standard deviations generated at different scales.

[0077] Furthermore, in step 5, the set of positive sample pairs... Specifically:

[0078]

[0079] Furthermore, in step 6, the negative sample pair set Specifically:

[0080]

[0081] in, negative sample pair set The k-th negative sample pair: including and Two simulated trajectory matching samples.

[0082] Furthermore, step 7 includes the following steps:

[0083] Step 701, set the positive sample pairs and negative sample pairs set The elements are merged, and in the new merged set, each element belongs to... or

[0084] Step 702: Each element obtains a corresponding simulated trajectory training matching sample pair based on its own contained trajectory information.

[0085]

[0086]

[0087] in, and They represent Trajectory information of two simulated trajectory matching samples in the corresponding elements; express The tag of the corresponding element, if The corresponding element belongs to but like The corresponding element belongs to but and They represent The longitude, latitude, and instantaneous heading of the first trajectory point corresponding to the simulated trajectory matching sample. and They represent The longitude, latitude, and instantaneous heading of the first trajectory point in the corresponding simulation trajectory matching sample, where n represents... and The number of trajectory points corresponding to the simulated trajectory matching samples, the value of n is determined by the set of simulated trajectories. The sum of the time intervals corresponding to the three trajectories and the sampling time interval Δt are determined;

[0088] Step 703 finally yields the set of training matching sample pairs for the simulated trajectory.

[0089]

[0090] Furthermore, in step 8, the distance feature set of the simulated trajectory training matching sample pairs is... Specifically:

[0091]

[0092] in, Representing the distance feature set The distance features of the training matching sample pair of the l-th simulated trajectory are calculated. hist() represents solving the histogram statistical frequency, and log() represents calculating the logarithm.

[0093] Furthermore, in step 9, the simulated trajectory training matches the direction feature set of the sample pairs. Specifically:

[0094]

[0095] in, Representation of directional feature set The directional features of the training matching sample pair of the l-th simulated trajectory.

[0096] Furthermore, the trajectory F to be matched in step 11 re and S re Specifically:

[0097] F re ={x F (1),y F (1),w F (1)),(x F (2),y F (2),w F (2)),...,

[0098] (x F (n),y F (n),w F (n))}

[0099] S re ={x S (1),y S (1),w S (1)),(x S (2),y S (2),w S (2)),...,

[0100] (x S (n),y S (n),w S (n))}

[0101] Where, x F (1) y F (1) and w F (1) Represent the trajectory F respectively re The longitude, latitude, and instantaneous heading of the first trajectory point, x S (1) yS (1) and w S (1) Represent the simulation trajectory S respectively re The longitude, latitude, and instantaneous heading of the first trajectory point, and the simulated trajectory F. re and S re The number of trajectory points is n.

[0102] In summary, this invention can effectively utilize the target motion element information in the trajectory to construct a matching model and achieve target trajectory matching; the trajectory matching results are more accurate and reliable, achieving the matching of the most similar path target trajectory, thereby providing strong technical support for the optimization and management of marine transportation.

[0103] Those skilled in the art will recognize that the described embodiments are intended to help readers understand the principles of the invention and should be understood as not limiting the scope of protection of the invention to the described embodiments. Various modifications and variations can be made to the invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the invention should be included within the scope of the claims of the invention.

Claims

1. A trajectory matching method combining joint discrimination of target motion features, characterized in that, Comprising the steps of: Step 1, for trajectory data of different densities in the simulation area R, set three sub-areas of different scales, denoted as R = {r1, r2, r3}; Step 2, randomly generate m trajectory starting points in each sub-region of the simulation region R, and construct a trajectory starting point set i = 1, 2, 3; m > 7; Step 3, generating a simulation trajectory set according to the trajectory starting point set generating a simulation trajectory set according to the trajectory simulation parameters Step 4, for each simulation trajectory in the simulation trajectory set two scale random noises are introduced respectively, so as to obtain two simulation trajectory matching samples of different scales, and record them as a simulation trajectory matching sample pair, and finally generate a simulation trajectory matching sample pair set Step 5, match all simulation trajectory pairs in the simulation trajectory pair set with the positive sample pair set to obtain a positive sample pair set Step 6, match the simulation trajectory matching sample pair set with the simulation trajectory matching sample pair set Cross-match any two different simulation trajectory matching sample pairs to obtain a plurality of simulation trajectory matching sample pairs not belonging to the simulation trajectory matching sample pair set From any 3m simulation trajectory matching sample pairs, mark the negative sample pairs as negative sample pairs, and give a label of 0 to obtain a negative sample pair set Step 7, Resample positive and negative sample pairs sets to construct simulated trajectory training matching sample pairs sets ​ Step 8, calculate distance features of each simulation trajectory training matching sample pair to obtain a distance feature set Step 8, calculate distance features of each simulation trajectory training matching sample pair to obtain a distance feature set Step 9, calculating the simulation trajectory training matching sample pair set obtaining a direction feature set by calculating the direction feature of each simulation trajectory training matching sample pair Step 10, input the distance feature set and the direction feature set of the simulation trajectory training matching sample pair set into the SVM classifier to build a trajectory matching classification model M SVM The input of the trajectory matching classification model M SVM is the distance feature and the direction feature between two trajectories, and the output is the label 0 or 1.​​​ Step 11, for the two trajectories F re and S re to be matched, the distance feature and the direction feature between the two are calculated, and then the distance feature and the direction feature are input into the trajectory matching classification model M SVM , to obtain the corresponding labels of the trajectories F re and S re . If the label is 0, it indicates that the trajectories F re and S re are not matched, and if the label is 1, it indicates that the trajectories F re and S re are matched, thereby completing the trajectory matching based on the joint discrimination of the target motion features.

2. The trajectory matching method of claim 1, wherein, The set of simulated trajectories in step 3 may be represented as: wherein, denotes a set of simulation trajectories denotes the jth simulation trajectory in the set of simulation trajectories: comprising three trajectory parameters and a sampling time interval, and denote the time interval corresponding to the three trajectory segments, respectively, and the sum of the time intervals corresponding to the three trajectory segments is the same for different simulation trajectories; and denote the initial velocity of the corresponding time interval, and denote the initial direction of the corresponding time interval, and denote the unit time velocity variable of the corresponding time interval, and denote the unit time direction variable of the corresponding time interval; Δt denotes the sampling time interval of the trajectory point, and the sampling time interval of the trajectory point is the same for different simulation trajectories.

3. The trajectory matching method of claim 2, wherein, The set of sample pairs of simulated trajectories in step 4 Specifically: wherein, a set of pairs of simulated trajectory matching samples a j-th simulated trajectory matching sample pair in the set and both are based on simulated trajectories correspondingly generated by adding position noise and respectively represent simulated trajectory matching samples generated different scale noise mean, and respectively represent simulated trajectory matching samples generated different scale standard deviation.

4. The trajectory matching method of claim 3, wherein, The positive sample pair set in step 5 Specifically:

5. The trajectory matching method of claim 4, wherein, The set of negative sample pairs in step 6 Specifically: wherein, is a set of negative sample pairs is the kth negative sample pair of the set of negative sample pairs: comprising and two simulated trajectory matching samples.

6. The trajectory matching method of claim 5, wherein, Step 7 comprises the steps of: Step 701, merge the positive sample pair set and the negative sample pair set , and each element in the new set after merging belongs to or Step 702, each element obtains the corresponding simulation trajectory training matching sample pair according to the trajectory information contained in itself wherein, and respectively represent the trajectory information of two simulated trajectory matching samples in the corresponding element; represent the label of the corresponding element, if the corresponding element belongs to then if the corresponding element belongs to then and respectively represent the longitude, latitude and instantaneous heading of the first trajectory point of the corresponding simulated trajectory matching sample, and respectively represent the longitude, latitude and instantaneous heading of the first trajectory point of the corresponding simulated trajectory matching sample, n represents and the number of trajectory points of the corresponding simulated trajectory matching sample, the value of n is determined by the sum of the time intervals corresponding to the three trajectories in the simulated trajectory set and the sampling time interval Δt. Step 703, finally obtaining the simulation trajectory training matching sample pair set 7. The trajectory matching method of claim 6, wherein, The distance feature set of the matched sample pair in step 8 is simulated trajectory training Specifically: wherein, representing distance features of the set of distance features distance feature of the l-th simulated trajectory training matching sample pair in the set of distance features, hist() represents a histogram frequency, and log() represents a logarithm.

8. The trajectory matching method of claim 7, wherein, The direction feature set of the matching sample pair in step 9 is trained by simulation trajectory Specifically: wherein, representing a set of directional features the directional features of the l-th simulated trajectory training matching sample pair in 9. The trajectory matching method of claim 8, wherein, The trajectories F to be matched in step 11 re and S re In particular: F re = {x F (1), y F (1), w F (1)), (x F (2), y F (2), w F (2)),..., (x F (n),y F (n),w F (n))} S re = {x S (1),y S (1),w S (1)), (x S (2),y S (2),w S (2)),..., (x S (n),y S (n),w S (n))} wherein x F (1), y F (1) and w F (1) respectively represent the longitude, latitude and instantaneous heading of the first trajectory point of the trajectory F re (1), y S (1) and w S (1) respectively represent the longitude, latitude and instantaneous heading of the first trajectory point of the trajectory F S (1), y re (1) and w re (1) respectively represent the longitude, latitude and instantaneous heading of the first trajectory point of the trajectory F re and S

Images