CN113344039A - Multi-extension target tracking method based on space-time correlation - Google Patents
Multi-extension target tracking method based on space-time correlation Download PDFInfo
- Publication number
- CN113344039A CN113344039A CN202110547305.4A CN202110547305A CN113344039A CN 113344039 A CN113344039 A CN 113344039A CN 202110547305 A CN202110547305 A CN 202110547305A CN 113344039 A CN113344039 A CN 113344039A
- Authority
- CN
- China
- Prior art keywords
- target
- matrix
- clustering
- gaussian
- extended
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 49
- 238000005259 measurement Methods 0.000 claims abstract description 31
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 21
- 239000011159 matrix material Substances 0.000 claims description 34
- 238000001914 filtration Methods 0.000 claims description 8
- 238000001514 detection method Methods 0.000 claims description 5
- 230000003595 spectral effect Effects 0.000 claims description 4
- 239000000126 substance Substances 0.000 claims description 4
- 230000007704 transition Effects 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 3
- 238000013138 pruning Methods 0.000 claims description 3
- 238000013459 approach Methods 0.000 claims description 2
- 238000010276 construction Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 claims description 2
- 238000007781 pre-processing Methods 0.000 claims description 2
- 238000005070 sampling Methods 0.000 claims description 2
- 238000000638 solvent extraction Methods 0.000 abstract description 3
- 101001086191 Borrelia burgdorferi Outer surface protein A Proteins 0.000 description 6
- 230000004083 survival effect Effects 0.000 description 5
- 238000005192 partition Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/15—Correlation function computation including computation of convolution operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention relates to a multi-extension target tracking method based on space-time correlation. When a target occupies multiple fractional rate units of the sensor, a single target will produce multiple measurements, i.e., an extended target. Under the background, when extended targets are crossed, a general distance-based partitioning method will classify measured values of different targets into the same measurement set, which causes the precision of the filter to be reduced and potential estimation to be wrong. The method is based on an ET-GM-PHD algorithm, adopts a space-time correlation idea, utilizes the correlation of measurement values of the extended targets at adjacent moments, and tracks the multiple extended targets on the basis of SNN division of a directed graph. The method greatly reduces the tracking error of the extended target at the intersection, and realizes accurate estimation on the number of the targets and the position of the targets. Meanwhile, the tracking process of the extended target and the point target is separated, and the calculated amount is greatly reduced.
Description
Technical Field
The invention belongs to the field of information fusion, and relates to a multi-extension target tracking method based on space-time correlation.
Background
With the development of modern sensor technology, the resolution of the sensor is higher and higher, a single target occupies multiple resolution units of the sensor, and multiple measurement information about the target is obtained at the same time, and the target type becomes an extended target. ET-GM-PHD (extended Target Gaussian Mixture PHD) is a filtering algorithm for tracking multiple extended targets under the theory of random finite set, and the method uses a Gaussian Mixture form to represent the intensity function of multiple targets so as to approximate the posterior distribution of the multiple targets. The multi-extended target method based on the random finite set avoids data association and reduces the complexity of calculation, wherein the indispensable part is to reasonably divide the measurement set at the current moment so as to find out the correct division of the measurement set. The widely used dividing method at present is distance division, and the method divides the measuring values meeting the distance threshold into the same dividing unit by utilizing the characteristic that the measuring values belonging to the same target are close to each other in space, so as to realize the division of the measuring set. However, when the target distance is close, the measurement values of different targets may fall into the same measurement set, resulting in filtering errors.
Disclosure of Invention
The invention aims to overcome the defects of the existing method, provides a multi-extension target tracking method based on space-time association, solves the problems of target state and target number estimation under the condition that the multi-extension target is closer in distance and the track is crossed, has good performance, and is not influenced by the measurement density. The method is mainly applied to the scene that a single sensor realizes the tracking of multiple extended targets in a clutter environment.
The technical scheme of the invention is as follows:
a multi-extended target tracking method based on space-time considers the relevance between measurement values of extended targets at adjacent moments, and specifically comprises the following steps:
by a single sensorCompleting the detection of the extended target, and making the measurement set obtained by the sensor at the moment k as ziTo represent a single measurement in two dimensions, nkIs the number of measured values; the extended target state set in the detection area of the sensor at the moment k isWherein M iskWhich indicates the number of the extension target,representing the state of motion, x, of a single extended objecti,yiThe information of the object is represented by,representing object motion information.
The motion equation of the extended target is established as follows:
Xk=FXk-1+vk
wherein the content of the first and second substances,is a state transition matrix, I is an identity matrix, TsIs the sampling interval. v. ofkRepresents a covariance ofProcess noise of σvIs the process standard deviation.
The measurement equation of the extended target is:
Zk=HXk+wk
wherein the content of the first and second substances,to observe the equation, wkRepresents a covariance ofOf the measured noise, σεTo measure the standard deviation of the noise.
A multi-extension target tracking method based on space-time association is shown in figure 1 and mainly comprises the following steps:
s1, when the time k is 0, the gaussian component in the initialized system isWherein w0Is the weight of the Gaussian component, m0Is the mean value of the Gaussian components, expressed as the motion state, P0For the corresponding covariance matrix, J0Is the initial number of gaussians;
s2, when k is larger than or equal to 1, traversing the Gaussian component set, performing one-step prediction according to the state transition matrix, and adding the Gaussian component of the newly generated target, wherein the steps are expressed as follows:
in the formula, Jk-1The number of gaussian components at time k-1,andrespectively, parameters of the newly generated target Gaussian components, JBRepresenting the number of new target gaussian components;
s3, selecting components with weight more than 0.5 in the Gaussian components:in the formula JcIs the number of gaussian components; and selects the position component thereofConstructing a wave gate, and setting a distance threshold as tau; traverse the entire measurement set ZkForming measurement sets for measurement values falling into the wave gatesN represents the number of metrology sets;
s4, calculating a measurement setUnion ofClustering by adopting a K-means + + algorithm to obtain L clustering centersClustering centers by adopting GM-PHD algorithm of point targetsUpdating the Gaussian component G to obtain an updated Gaussian component:in the formulaIs the number of gaussian components;
s5, calculating a k time measurement set ZkAnd collectionsDifference set ofClustering set by DBSCAN clustering algorithmPreprocessing, if a clustering unit exists at the moment, turning to S6; otherwise, go to S7;
s6, for the clustering unit generated in S5, calculating the union of the clustering unitsCarry out one by utilizing a directed graph SNN clustering methodStep processing, the clustering number of the step processing is K e to Kl,Ku]In which K islFor the number of clusters in the DBSCAN algorithm,n is a setThe number of the measured values, beta, is the measurement rate of the extended target;
s61 construction setSimilarity matrix W of directed kNN adjacency matrixn×nSelecting k as 8 in the algorithm; the elements in W are:
wherein z issIs ziTo (1) aA nearest neighbor point of zjTo (1) aA nearest neighbor point, NiDenotes ziK nearest neighbors, N, in a kNN graphjDenotes zjK nearest neighbors, z, in a kNN graphi、zjAnd zsRespectively being vertexes in the kNN graph; n is a radical ofi∩NjDenotes ziAnd zjThe shared closest point of approach between, defined as:
in the formula, zijIs a virtual vertex, representing a vertex ziAnd zjAre the closest points to each other.
S62 matrix W andcluster number range [ K ]l,Ku]Clustering by adopting a spectral clustering method to form a clustering unit; the spectral clustering comprises the following steps:
(1) calculating an unnormalized laplacian matrix L from the similarity matrix W,
in the formula, D is a diagonal matrix and is a degree matrix of the kNN diagram.
(2) Computing a normalized Laplace matrix Ls=D-1/2LD-1/2;
(3) Calculating the matrix LsCharacteristic value λ ofiAnd a feature vector ui;
(4) Selecting eigenvectors corresponding to K minimum eigenvalues to form a matrix U;
(5) and normalizing the matrix U to obtain a matrix Y, taking each row of the matrix Y as a new data point of a K-dimensional space, and clustering by using a K-means + + algorithm.
S7, on the basis of the S6 division result, updating the Gaussian component G by utilizing an ET-GM-PHD algorithm, and obtaining the updated Gaussian component as follows: is the number of gaussian components.
S8, for Gaussian componentPruning and merging are carried out, the Gaussian components with the weight smaller than T are deleted, and the Gaussian components with the distance smaller than mu between the merged Gaussian components are deleted.
And S9, selecting the Gaussian component with the weight greater than 0.5 as a filtering result of the target, and realizing target tracking.
The invention has the beneficial effects that: the method reduces the calculation amount of filtering, and can effectively process the target position estimation and the target number estimation of the extended target at the intersection. The extended target tracking process is converted into the form of extended target tracking and point target tracking, the operation efficiency is improved, meanwhile, the filtering algorithm has better robustness, and the interference of clutter on the estimation result is reduced.
Drawings
FIG. 1 is a flow chart of the present invention as applied to multi-extended target tracking;
FIG. 2 shows the flight path estimation, the real flight path and the measured value in the embodiment 1;
FIG. 3 OSPA error comparison under example 1;
FIG. 4 comparison of the number of targets estimated in example 1;
FIG. 5 run time comparison under example 1;
FIG. 6 shows the flight path estimation, the real flight path, and the measured value in embodiment 2;
FIG. 7 OSPA error comparison under example 2;
FIG. 8 target number estimation in example 2;
FIG. 9 run time comparison under example 2;
Detailed Description
The simulation parameters are set as follows: target survival probability of Ps0.99, with a probability of detection of PDWhen the number of the extended target measurement is 0.99, the number of the extended target measurement follows Poisson distribution with the expected value of 10; the clutter measurement number follows Poisson distribution with the expected value of 20; the maximum number of Gaussian components is J max100, the pruning threshold T is 10-6The merging threshold mu is 4; the distance threshold τ is 50. Covariance of process noise of s v2; observing a variance of noise ofsεThe monitored area size is 20: [ -1000, 1000]×[-1000,1000]m。
Examples 1,
The purpose of the embodiment is to verify that the proposed method is at two extended targetsAnd the filtering performance of the target under the scene of close distance and parallel motion. The initial state of the target 1 is [ -600m, -500m,0m/s]TThe survival time is 1-100 s; the initial state of the target 2 is [ -600m, -600m,0m/s]TThe survival time is 1-100 s; the intensity function of the nascent object in this scenario is:
FIG. 2 is a comparison of the target state estimation and target true state of the present invention under example 1. It can be seen that the invention can obtain a better tracking result even when the extended targets are closer to each other.
FIG. 3 is a comparison of OSPA errors between the method of the present invention and the distance-and directed-graph SNN-partitioning method in example 1. It can be seen that the OSPA error of the method of the present invention is significantly minimized.
Fig. 4 is a comparison of the target number estimation of the method of the present invention and the distance-division and directed graph SNN-division method in embodiment 1. It can be seen that the target number estimation of the method of the present invention is most accurate.
Fig. 5 is a comparison of the algorithm running time of the method of the present invention and the distance partition and directed graph SNN partition method in embodiment 1, and it can be seen that the algorithm running time of the method of the present invention is the lowest.
Examples 2,
The simulation parameters were the same as in example 1. The purpose of this embodiment is to verify the filtering performance of the algorithm in case of target crossing. 2 targets exist in the scene, and the initial state of the target 1 is [250m,250m,0m/s]TThe survival time is 1-100 s; the initial state of the target 2 is [ -250m, -250m,0m/s]TThe survival time is 1-100 s; the two objects intersect at 56 s. The intensity function of the nascent object in this scenario is:
FIG. 6 is a comparison of the target state estimation and the target true state of the present invention in example 2. It can be seen that a better estimation result can be obtained even at the target track crossing time.
FIG. 7 shows the OSPA error comparison of the method of the present invention with the distance-and directed-graph SNN-partitioning method in example 2. It can be seen that the OSPA error of the method of the present invention is minimal even at extended target track crossing times.
Fig. 8 is a comparison of the target number estimation of the method of the present invention and the distance-division and directed graph SNN-division method in embodiment 2. It can be seen that the target number estimation of the method is the most accurate at the track crossing time.
Fig. 9 is a comparison of the algorithm running time of the method of the present invention and the distance partition and directed graph SNN partition method in embodiment 2, and it can be seen that the algorithm running time of the method of the present invention is the lowest.
Claims (1)
1. A space-time-based multi-extension target tracking method is characterized by comprising the following steps:
the detection of the extended target is finished by a single sensor, and the measurement set obtained by the sensor at the moment k is combined intoziTo represent a single measurement in two dimensions, nkIs the number of measured values; the extended target state set in the detection area of the sensor at the moment k isWherein M iskWhich indicates the number of the extension target,representing the state of motion, x, of a single extended objecti,yiThe information of the object is represented by,representing target motion information;
the motion equation of the extended target is established as follows:
Xk=FXk-1+vk
wherein the content of the first and second substances,is a state transition matrix, I is an identity matrix, TsIs the sampling interval; v. ofkRepresents a covariance ofProcess noise of σvIs the process standard deviation;
the measurement equation of the extended target is:
Zk=HXk+wk
wherein the content of the first and second substances,to observe the equation, wkRepresents a covariance ofOf the measured noise, σεMeasuring the standard deviation of the noise;
the method for tracking the extended target comprises the following specific steps:
s1, when the time k is 0, the gaussian component in the initialized system isWherein w0Is the weight of the Gaussian component, m0Is the mean value of the Gaussian components, expressed as the motion state, P0For the corresponding covariance matrix, J0Is the initial number of gaussians;
s2, when k is larger than or equal to 1, traversing the Gaussian component set, performing one-step prediction according to the state transition matrix, and adding the Gaussian component of the newly generated target, wherein the steps are expressed as follows:
in the formula, Jk-1The number of gaussian components at time k-1,andrespectively, parameters of the newly generated target Gaussian components, JBRepresenting the number of new target gaussian components;
s3, selecting components with weight more than 0.5 in the Gaussian components:in the formula JcIs the number of gaussian components; and selects the position component thereofConstructing a wave gate, and setting a distance threshold as tau; traverse the entire measurement set ZkForming measurement sets for measurement values falling into the wave gatesN represents the number of metrology sets;
s4, calculating a measurement setUnion ofClustering by adopting a K-means + + algorithm to obtain L clustering centersClustering centers by adopting GM-PHD algorithm of point targetsUpdating the Gaussian component G to obtain an updated Gaussian component:in the formulaIs the number of gaussian components;
s5, calculating a k time measurement set ZkAnd collectionsDifference set ofClustering set by DBSCAN clustering algorithmPreprocessing, if a clustering unit exists at the moment, turning to S6; otherwise, go to S7;
s6, for the clustering unit generated in S5, calculating the union of the clustering unitsFurther processing by utilizing a directed graph SNN clustering method, wherein the clustering number is K-E [ K ∈l,Ku]In which K islFor the number of clusters in the DBSCAN algorithm,n is a setThe number of the measured values, beta, is the measurement rate of the extended target;
wherein z issIs ziTo (1) aA nearest neighbor point of zjTo (1) aA nearest neighbor point, NiDenotes ziK nearest neighbors, N, in a kNN graphjDenotes zjK nearest neighbors, z, in a kNN graphi、zjAnd zsRespectively being vertexes in the kNN graph; n is a radical ofi∩NjDenotes ziAnd zjThe shared closest point of approach between, defined as:
in the formula, zijIs a virtual vertex, representing a vertex ziAnd zjAre the closest points to each other;
s62, for similarity matrix W and cluster number range [ Kl,Ku]Clustering by adopting a spectral clustering method to form a clustering unit; the spectral clustering comprises the following steps:
(1) calculating an unnormalized laplacian matrix L from the similarity matrix W,
in the formula, D is a diagonal matrix and represents a degree matrix of the kNN diagram;
(2) computing a normalized Laplace matrix Ls=D-1/2LD-1/2;
(3) Calculating the matrix LsCharacteristic value λ ofiAnd a feature vector ui;
(4) Selecting eigenvectors corresponding to K minimum eigenvalues to form a matrix U;
(5) and normalizing the matrix U to obtain a matrix Y, taking each row of the matrix Y as a new data point of a K-dimensional space, and clustering by using a K-means + + algorithm.
S7, on the basis of the S6 division result, updating the Gaussian component G by utilizing an ET-GM-PHD algorithm, and obtaining the updated Gaussian component as follows: is the number of gaussian components;
s8, for Gaussian componentPruning and combining are carried out, the Gaussian components with the weight smaller than T are deleted, and the Gaussian components with the distance smaller than mu between the combined Gaussian components are deleted;
and S9, selecting the Gaussian component with the weight larger than 0.5 as the filtering result of the target.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110547305.4A CN113344039B (en) | 2021-05-19 | 2021-05-19 | Multi-extension target tracking method based on space-time correlation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110547305.4A CN113344039B (en) | 2021-05-19 | 2021-05-19 | Multi-extension target tracking method based on space-time correlation |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113344039A true CN113344039A (en) | 2021-09-03 |
CN113344039B CN113344039B (en) | 2022-04-26 |
Family
ID=77469540
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110547305.4A Expired - Fee Related CN113344039B (en) | 2021-05-19 | 2021-05-19 | Multi-extension target tracking method based on space-time correlation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113344039B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116256746A (en) * | 2023-05-12 | 2023-06-13 | 江苏亿连通信技术有限公司 | Radar-based system and method for monitoring intrusion of foreign matters into perimeter airspace of preventive area |
CN117214857A (en) * | 2023-11-09 | 2023-12-12 | 中国人民解放军海军航空大学 | Tracking method of Gaussian multi-hypothesis multi-expansion target in three-dimensional scene |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109671096A (en) * | 2017-10-13 | 2019-04-23 | 南京航空航天大学 | A kind of space-time neighbour target detection and Grid Clustering measure more extension method for tracking target under dividing |
CN109991570A (en) * | 2019-04-05 | 2019-07-09 | 电子科技大学 | A kind of direct localization method of moving target based on quasi-Newton method |
CN110297221A (en) * | 2019-06-19 | 2019-10-01 | 西安电子科技大学 | A kind of data correlation method based on gauss hybrid models |
CN110298865A (en) * | 2019-05-22 | 2019-10-01 | 西华大学 | The space-based Celestial Background small point target tracking of cluster device is separated based on threshold value |
CN110992396A (en) * | 2019-10-18 | 2020-04-10 | 河南科技大学 | K-means + + clustering algorithm-based multi-extension target tracking method |
US20200300972A1 (en) * | 2015-07-17 | 2020-09-24 | Origin Wireless, Inc. | Method, apparatus, and system for vital signs monitoring using high frequency wireless signals |
-
2021
- 2021-05-19 CN CN202110547305.4A patent/CN113344039B/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20200300972A1 (en) * | 2015-07-17 | 2020-09-24 | Origin Wireless, Inc. | Method, apparatus, and system for vital signs monitoring using high frequency wireless signals |
CN109671096A (en) * | 2017-10-13 | 2019-04-23 | 南京航空航天大学 | A kind of space-time neighbour target detection and Grid Clustering measure more extension method for tracking target under dividing |
CN109991570A (en) * | 2019-04-05 | 2019-07-09 | 电子科技大学 | A kind of direct localization method of moving target based on quasi-Newton method |
CN110298865A (en) * | 2019-05-22 | 2019-10-01 | 西华大学 | The space-based Celestial Background small point target tracking of cluster device is separated based on threshold value |
CN110297221A (en) * | 2019-06-19 | 2019-10-01 | 西安电子科技大学 | A kind of data correlation method based on gauss hybrid models |
CN110992396A (en) * | 2019-10-18 | 2020-04-10 | 河南科技大学 | K-means + + clustering algorithm-based multi-extension target tracking method |
Non-Patent Citations (2)
Title |
---|
KARL GRANSTROM等: "Extended Target Tracking using a Gaussian-Mixture PHD Filter", 《IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS》 * |
蒲磊等: "基于SNN-AP 聚类的扩展目标量测集划分方法", 《计算机应用研究》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116256746A (en) * | 2023-05-12 | 2023-06-13 | 江苏亿连通信技术有限公司 | Radar-based system and method for monitoring intrusion of foreign matters into perimeter airspace of preventive area |
CN117214857A (en) * | 2023-11-09 | 2023-12-12 | 中国人民解放军海军航空大学 | Tracking method of Gaussian multi-hypothesis multi-expansion target in three-dimensional scene |
CN117214857B (en) * | 2023-11-09 | 2024-02-02 | 中国人民解放军海军航空大学 | Tracking method of Gaussian multi-hypothesis multi-expansion target in three-dimensional scene |
Also Published As
Publication number | Publication date |
---|---|
CN113344039B (en) | 2022-04-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113344039B (en) | Multi-extension target tracking method based on space-time correlation | |
CN107831490A (en) | A kind of improved more extension method for tracking target | |
CN110503071B (en) | Multi-target tracking method based on variational Bayesian label multi-Bernoulli superposition model | |
CN110049549B (en) | WiFi fingerprint-based multi-fusion indoor positioning method and system | |
CN111652070B (en) | Face sequence collaborative recognition method based on monitoring video | |
CN112508963B (en) | SAR image segmentation method based on fuzzy C-means clustering | |
Qin et al. | Application of an efficient graph-based partitioning algorithm for extended target tracking using GM-PHD filter | |
CN107832778A (en) | A kind of same target recognition methods based on spatial synthesis similarity | |
Cao et al. | Correlation-based tracking of multiple targets with hierarchical layered structure | |
CN106780567B (en) | Immune particle filter extension target tracking method fusing color histogram and gradient histogram | |
CN111488552B (en) | Close-proximity multi-target tracking method based on Gaussian mixture probability hypothesis density | |
CN114002667A (en) | Multi-neighbor extended target tracking algorithm based on random matrix method | |
Leung et al. | Evaluating set measurement likelihoods in random-finite-set slam | |
CN108009272B (en) | Low-altitude moving target searching method based on directed weighted graph | |
CN115619825A (en) | Ground multi-target tracking state and track determining method | |
CN111273277B (en) | Prediction-guided multi-extension target measurement set partitioning method | |
Tian et al. | Partition of GB-InSAR deformation map based on dynamic time warping and $ k $-means | |
Yuan et al. | Research on face tracking algorithm based on detection and supervision tracking | |
Chen et al. | A mean shift algorithm based on modified Parzen window for small target tracking | |
CN112241002A (en) | Novel robust closed-loop detection method based on Karto SLAM | |
Xinglin et al. | Fast density peak-based clustering algorithm for multiple extended target tracking | |
Li et al. | Extended Target Measurement Set Partition Algorithm Based on KFCM Clustering | |
Han et al. | An Extended Target Tracking Algorithm Based on Gaussian Surface Fitting | |
Jiang et al. | Marine Targets Spatiotemporal Trajectory Similarity Measurement Method Based on Multidimensional Features | |
Chen et al. | A Samplable Multimodal Observation Model for Global Localization and Kidnapping |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20220426 |
|
CF01 | Termination of patent right due to non-payment of annual fee |