CN109814074B - Group target tracking method based on image morphological processing - Google Patents
Group target tracking method based on image morphological processing Download PDFInfo
- Publication number
- CN109814074B CN109814074B CN201910057384.3A CN201910057384A CN109814074B CN 109814074 B CN109814074 B CN 109814074B CN 201910057384 A CN201910057384 A CN 201910057384A CN 109814074 B CN109814074 B CN 109814074B
- Authority
- CN
- China
- Prior art keywords
- group
- target
- center
- targets
- sampling period
- 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.)
- Active
Links
Images
Landscapes
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a group target tracking method based on image morphological processing, which comprises the steps of processing echo signals received by a radar receiver in a sampling period, carrying out coordinate conversion calculation, establishing an initial image matrix and calculating a binary image matrix, carrying out image morphological processing and marking processing on the binary image matrix in sequence, calculating an equivalent center vector of a group target in the sampling period, calculating a track set of the group target by adopting a logic method, calculating an initial state of the center of the group target, calculating an association matrix, adjusting a filtering tracking track according to the association matrix, taking the current filtering tracking track state as a filtering state of the center position prediction of the group target in the next sampling period, and finally obtaining a group target tracking result. The method of the invention is used in the technical field of radar tracking, not only ensures the calculation precision of the group target center, but also well tracks the group target under the condition that part of the group target is lost, and has valuable practicability.
Description
Technical Field
The invention belongs to the technical field of radar target tracking, and particularly relates to a group target tracking method based on image morphological processing.
Background
In modern war, the detection and attack by unmanned aerial vehicle technology has become an important means for military countermeasures, and the threat of unmanned aerial vehicle group combat on defense systems is very huge. The group target can be regarded as a group of targets which are close to each other and have basically consistent speed and movement direction in a short time. Therefore, the center of the group target is tracked to maintain the overall tracking of the group target, and the position of each group in each frame is obtained by tracking the center of the group target without considering each target in the group; the group target tracking has important application value in the field of aerial target monitoring and unmanned aerial vehicle multi-target formation motion tracking. Since the radar resolution unit is affected by the beam width, the occurrence of dense group targets will affect the accuracy of detecting the echoes, resulting in group target observations obtained by the radar that are partially resolvable. The reason why the two targets are partially distinguishable is that the distance and angle between the targets in the same cluster exceed the radar resolution capability, and the situation that two targets are recognized as one target often occurs, so that the observation information of all the single targets in the cluster cannot be obtained. This increases the position error between the center of the cluster target and the true center, which may lead to missing cluster targets in severe cases.
Disclosure of Invention
In order to solve the technical problem, the invention discloses a group target tracking method based on image form processing, which is characterized in that the method comprises the steps of calculating the number of group targets and the coordinates of a group target equivalent center in a sampling period by using image morphology and a connected region marking method based on an observed value obtained by processing an echo received by a radar receiver in the sampling period, further using a logic method to carry out inference calculation on a track set of the group target center, determining the initial state of the group target center, then predicting the group target center position in the next sampling period according to the initial state of the group target center, associating the group target center vector in the next sampling period with the corresponding group target equivalent center position vector in the next sampling period, and finally adjusting a filtering tracking parameter according to an association result to further track the group target. The method improves the accuracy of the radar point trace preprocessing module in calculating the group target center based on the image morphological processing, reduces the complexity of subsequent tracking, and improves the efficiency of tracking the group target center when the radar data processing module adopts a Kalman filtering algorithm.
In order to achieve the purpose, the invention adopts the following technical scheme to realize the purpose:
a group target tracking method based on image morphological processing comprises the following steps:
And 2, establishing an initial image matrix I and calculating a binary image matrix I'.
And 3, sequentially carrying out image morphological processing and marking processing on the binary image matrix I' to obtain the number L of group targets and an image matrix Ipm, L =1, 2.
Step 4, calculating equivalent central vectors Z, Z = (Z) of L group targets in the sampling period according to the image matrix Ipm containing different mark numbers L 1 ,z 2 ,…z l ,…,z L ) Wherein z is l =(Xc l ,Yc l ),Xc l ,Yc l Respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the ith group target in the rectangular coordinate system in the sampling period.
Step 6, calculating the initial states of the L group target centers, wherein the initial state of the ith group target center is as follows:
wherein, X _ st (l) Indicating the initial state of the ith group object, X _ initial (l) 、Y_initial (l)、 X-direction coordinates and Y-direction coordinates respectively representing the center of the first group target in the initial state, VX _ initial (l) 、VY_initial (l) Respectively representing the l group target centersX-direction velocity and Y-direction velocity in the initial state.
And 7, the radar data processing module starts to track the L group targets by taking the initial states of the centers of the L group targets obtained in the step 6 as a reference, the total tracking steps are set to be K, and K =0 is initialized.
Step 8, after receiving all echo signals in the next sampling period, the radar receiver executes the steps 1 to 4, so that equivalent central vectors Z of F group targets in the next sampling period are obtained c Then Z is c =(z 1 ,z 2 ,…z f ,…,z F )。
Step 9, equivalent central vectors Z of F group targets in the next sampling period c And the predicted central position vector Z corresponding to the L group targets in the next sampling period c Performing correlation calculation to obtain a correlation matrix C, and obtaining predicted central position vectors Z of the L group targets in the next sampling period c ' calculated in substep 3 of this step 9;
the step 9 specifically includes the following three substeps:
substep 2, when F > L, indicating that the tracked L group targets are separated, then continuing to execute step 4 from step 5 to step 6 to recalculate the initial states of the centers of the F group targets, and then executing substep 3 in step 9;
and a substep 3 of adding 1 to k and calculating the correlation matrix C.
and meanwhile, taking the filtering tracks of the current F group targets as the filtering states for predicting the central positions of the F group targets in a new next sampling period on one hand, and as the tracking results of the F group targets on the other hand.
And 11, judging the value of K, executing the steps 8 to 10 if K is less than K, and ending the tracking if K is less than K.
Compared with the prior art, the group target tracking method based on image form processing provided by the invention ensures that the deviation degree of the group target observation center can be reduced under the condition that the observation of the group target part is lost by utilizing the image form processing method, so that the obtained group target equivalent observation center is more accurate, and the tracking stability is improved; the method is applied to the technical field of radar tracking, the calculation accuracy of the radar point trace preprocessing module on the group target center is improved, and part of group targets with low resolution can be processed, so that the tracking accuracy of the group target center is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flowchart of a group target tracking method based on image morphological processing according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a group target image before processing according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a group target image after image morphological processing according to an embodiment of the invention;
FIG. 4 is a tracking trajectory diagram of the equivalent center of a group target based on image morphological processing according to an embodiment of the present invention;
FIG. 5 is a diagram of real trajectories observed at the center of a cluster target in accordance with an embodiment of the present invention;
FIG. 6 is a schematic diagram illustrating a distance error of a center track of a group target 1 according to an embodiment of the present invention;
FIG. 7 is a schematic diagram illustrating an azimuthal error of a center locus of a group target 1 according to an embodiment of the present invention;
FIG. 8 is a schematic diagram illustrating a distance error of a center track of a group target 2 according to an embodiment of the present invention;
FIG. 9 is a schematic diagram of an azimuth error of a center track of the group target 2 according to an embodiment of the present invention;
FIG. 10 is a schematic diagram illustrating a distance error of a center track of a group target 3 according to an embodiment of the present invention;
fig. 11 is a schematic view of an azimuth error of a center track of a group target 3 according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments, and all other embodiments obtained by a person skilled in the art without making creative efforts based on the embodiments of the present invention belong to the protection scope of the present invention.
Fig. 1 is a schematic flowchart illustrating a group target tracking method based on image morphological processing according to an embodiment of the present invention.
As shown in fig. 1, a group target tracking method based on image morphological processing according to an embodiment of the present invention includes the following steps:
firstly, sequentially carrying out pulse pressure, constant false alarm detection and monopulse angle measurement processing on all echo signals received by a radar receiver in a sampling period so as to obtain the total number Q of targets, the distance measured by each target and the azimuth angle measured by each target in all the echo signals received by the radar in the sampling period;
secondly, converting the distance and the azimuth angle measured by each target into an X-direction coordinate and a Y-direction coordinate of each target in a two-dimensional rectangular coordinate system by adopting a rectangular coordinate conversion method according to the following formula:
x d =R d cosθ d ,y d =R d sinθ d
wherein x is d 、y d Respectively representing the X-direction coordinate and the Y-direction coordinate of the d-th target measurement in a two-dimensional rectangular coordinate system, R d Represents the distance, θ, of the d-th target measurement d Indicating the azimuth angle of the d-th target measurement.
The step 1 is to perform correlation processing on echo signals received by the radar receiver in a sampling period so as to obtain the total number of actually observed targets and the position of each target under a polar coordinate, and lay a foundation for converting the position of each target under the polar coordinate into a rectangular coordinate and then applying image processing in a later period.
Step 2, establishing an initial image matrix I, and calculating a binary image matrix I', specifically comprising the following substeps:
Meanwhile, the size of a pixel unit of an image is x _ im × y _ im, wherein x _ im and y _ im respectively represent the width and the height of the pixel unit of the image.
And a substep 2 of calculating the dimensions of the rows and columns of the image matrix I, wherein the calculation formula is as follows:
where P denotes the dimension of a row of the image matrix I and N denotes the dimension of a column of the image matrix I.
And a substep 3 of initializing all pixel units in the image matrix I to an initial value of 0.
And step 4, calculating the X-direction coordinate and the Y-direction coordinate of all the target measurements in the image matrix I in the step 1, wherein the calculation formula is as follows:
wherein m =1,2, \8230, Q; x m _Im、Y m Im respectively represents the X-direction coordinate and the Y-direction coordinate of the mth target measurement in the image matrix I; x m _dkr、Y m The kr represents the X-direction coordinate and the Y-direction coordinate of the mth target measurement respectively;<>represents rounding down; x min represents the minimum value of the target monitoring area range in the X direction, and Y min represents the minimum value of the target monitoring area range in the Y direction.
And a substep 5, finding a coordinate position corresponding to each target measurement in the image matrix I according to the X-direction coordinate and the Y-direction coordinate of all the obtained target measurements in the image matrix I, and updating and assigning the initial value 0 of the corresponding pixel unit at the coordinate position to be 1, thereby obtaining a new image matrix I 'containing all the target measurements, and then the image matrix I' is a binary image matrix.
In the step 2, all the targets obtained in the step 1 are established in the image matrix I' through the rectangular coordinates of each target, so that preparation is made for processing the target tracking center of the group by using image morphology, and the method is also the key point of the method.
And 3, sequentially carrying out image morphological processing and marking processing on the binary image matrix I' to obtain the number L of group targets and an image matrix Ipm containing different marking numbers L, wherein L =1, 2., a, \8230andL, and the specific processes of the image morphological processing and the marking processing are as follows:
firstly, the binary image matrix I' is morphologically processed:
setting the structural element B as a matrix of a disk-shaped structure with the radius of 8 according to an image expansion algorithmPerforming expansion operation on the binary image matrix I', so as to obtain an edge-expanded binary image matrix Ip, wherein a plurality of connected regions are formed in the binary image matrix Ip; the dimensions of the rows and the columns in the binary image matrix Ip are respectively consistent with the dimensions of the rows and the columns in the image matrix I'.
Secondly, a connected domain marking method is adopted to perform marking calculation on a plurality of connected regions in the binary image matrix Ip, and the marking calculation process comprises the following steps:
3a) Scanning the binary image matrix Ip line by line in a line sequence, and if one or more pixel units in all pixel units in the first line are 1, executing 3 b) to 3 d); if all the pixel units in the first row are 0, continuing to scan the second row, and if all the pixel units in the second row are 0, continuing to scan the third row until the pixel unit in the scanned row appears 1 for the first time, and executing 3 b) to 3 d);
3b) Scanning a binary image matrix Ip line by line, forming a sequence by using continuous pixel units in each line as 1, and recording a starting point, an end point and a line number of each sequence;
for a row of which the pixel unit appears as 1 for the first time in all rows of the binary image matrix Ip, all sequences in the row are marked in sequence, and the corresponding mark number l is assigned as 1,2, \ 8230;, a' in sequence.
3c) Marking all sequences in the next line in sequence on the basis of the mark numbers of all sequence marks in the line of 3 b), wherein if the s-th sequence has a superposition area with one sequence in the line of 3 b), the mark number of the s-th sequence mark is the same as the mark number of the sequence mark in the line of 3 b); if the s-th sequence has an overlapping area with 3 b) two or more sequences in said row, the index number of the s-th sequence mark is the same as the smallest index number of the two or more sequence marks in 3 b) said row, and the index numbers of the two or more sequence marks in 3 b) said row are recorded as an equivalent pair; if the s-th sequence does not overlap with 3 b) all sequences in said row, the mark number of the s-th sequence mark is a new mark number, when s =1, the new mark number l = a' +1, when s ≠ 1, the new mark numbers are sorted according to the new mark numbers marked by said next row of sequences;
and by analogy, all sequences in all the following rows in the binary image matrix Ip are marked in sequence.
3d) The mark numbers of the different equivalent pairs recorded in 3 c) are respectively equivalent to the mark number of the minimum value in the mark numbers, the calculation for eliminating the equivalent pairs is completed, then all previous marks are traversed and all connected regions are re-marked in a natural number sequence, so that an image matrix Ipm containing different mark numbers L is obtained, the maximum value of the marks is the number L of the connected regions, the number of the connected regions is the number of the group targets, and the dimensions of the rows and the columns of the image matrix Ipm are respectively consistent with the dimensions of the rows and the columns in the image matrix Ip.
The step 3 is the most important of the method, the population number of all the targets in the step 1, namely the number of the group targets and the geometric area of each group target are found by using image morphology and a connected domain marking method, and the geometric area of each group target is marked by a marking number and is presented in an image matrix Ipm. Therefore, the center of the geometric area of each group of targets is fixed, and even if one or more targets in the group of targets are missing, the center of the geometric area of the group of targets is not influenced, so that the calculation accuracy of the center of each group of targets is ensured, and the tracking effect of the center of the group of targets is improved.
In the step 4, the step of,calculating equivalent central vectors Z, Z = (Z is) of L group targets in the sampling period according to the image matrix Ipm containing different mark numbers L 1 ,z 2 ,…z l ,…,z L ) Wherein z is l =(Xc l ,Yc l ),Xc l ,Yc l Respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the ith group target in the rectangular coordinate system in the sampling period:
as can be seen from step 3, the number of connected regions in the image matrix Ipm is L, and correspondingly, there are L group targets in all the targets received by the radar receiver in one sampling period.
First, the coordinates of the centers of all connected regions in the image matrix Ipm are calculated, and the calculation formula is as follows:
wherein the content of the first and second substances,the X-direction coordinate representing the center of the l-th connected component in the image matrix Ipm,a Y-direction coordinate representing the center of the l-th connected region in the image matrix Ipm; x is a radical of a fluorine atom ln ,y lj Respectively representing the X-direction coordinate of the nth pixel unit and the Y-direction coordinate of the jth pixel unit in the ith communication region; lr and lc respectively represent the total number of X-direction pixel units and the total number of Y-direction pixel units marked by l in the pixel unit in the image matrix Ipm; s l Is the total number of pixel cells in the l-th connected region.
Secondly, calculating the coordinates of the equivalent centers of the L group targets, wherein the calculation formula is as follows:
wherein, xc l ,Yc l Respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the ith group target in a rectangular coordinate system.
Finally, according to the coordinates of the equivalent centers of the L group targets, obtaining equivalent center vectors Z = (Z) of the L group targets in the sampling period 1 ,z 2 ,…z l ,…,z L ) Wherein z is l =(Xc l ,Yc l )。
And 4, obtaining the geometric area center of each group of targets by a method of solving the center of a mathematical geometric figure, and then converting the geometric area center of each group of targets into a rectangular coordinate system to obtain the equivalent center coordinate of each group of targets.
One sampling period is a sampling time, the sampling periods are sampling times, the radar sampling time is T, and preferably, the radar sampling time is the first four times, when T =1, T =2, T =3, and T =4, step 1 to step 4 are respectively performed, so as to obtain equivalent center vectors of L group targets at corresponding times: let T = T, T =1,2,3,4, the equivalent center vector of L group targets at the T-th time is Z t ,Z t =(z (t,1) ,z (t,2) ,…,z (t,L) ) Wherein z is (t,l) =(Xc t,l ,Yc t,l ),Xc t,l ,Yc t,l Respectively representing the X-direction of the equivalent center of the ith group target at the t moment in a rectangular coordinate systemTo the coordinates and Y-direction coordinates.
According to the embodiment of the invention, the sampling time is preferably determined by reasoning at the first four moments according to a logical method and simulating multiple times of experiments, and the sampling time is generally determined at the first three or four moments.
Therefore, the equivalent center coordinates of the L group targets at the first four moments are utilized, and the stable tracks of the L group targets are inferred by adopting a logic method:
firstly, the equivalent centers of L group targets in a first time are used for establishing an initial correlation wave gate by a speed method, and a possible track set is established for the equivalent centers of the L group targets in a second time falling into the initial correlation wave gate.
Secondly, extrapolating each subset track in the possible track set, taking an extrapolation point as a center, and determining the size of the wave gate at the moment by the covariance of track extrapolation errors: if the equivalent center of the group target falling into the correlation wave gate exists in the third moment, storing the equivalent center of the group target with the nearest extrapolation point into the corresponding possible track subset; if no equivalent center of the group target falls within the relevant wave gate, then the possible subset of tracks is deleted from the possible track set.
Finally, continuing to extrapolate the remaining possible flight paths, if a group equivalent center falling into a related wave gate exists in the fourth moment, storing the group equivalent center closest to the extrapolated point into the corresponding possible flight path set, and determining the flight path as a stable flight path; if no group equivalent center falls within the relevant wave gate, the possible track is deleted from the possible track set.
After the processing according to the logic method, a stable track set H of L group targets can be obtained, wherein H = { H = } 1 ,H 2 ,…,H l ,…,H L In which H l Is the stable track of the first group target and is a subset formed by the equivalent center coordinate position of the first four time first group target, then H l Is expressed as H l ={h (l,1) ,…,h (l,t) ,…h (l,4) In which the element h (l,t) Is the equivalent center coordinate position of the ith group target at the t-th time, i.e.Xc (l,t) 、Yc (l,t) Respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the ith group target in the rectangular coordinate system at the tth moment.
Step 6, calculating the initial state of L group target centers, wherein the initial state of the ith group target center is as follows:
wherein, X _ st (l) Represents the initial state of the ith cluster target center, X _ initial (l) 、Y_initial (l) Respectively representing the X-direction coordinate and the Y-direction coordinate of the center of the first group target in the initial state, VX _ initial (l) 、VY_initial (l) Respectively representing the speed of the ith group target center in the X direction and the speed of the ith group target center in the initial state.
The initial state of the ith cluster target center is calculated according to the following formula:
wherein Xc (l,T) 、Yc (l,T) Respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the first group target at the Tth time in a rectangular coordinate system, xc (l,t+1) 、Yc (l,t+1) Respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the ith group target at the t +1 th moment in a rectangular coordinate system, and delta t represents the time interval between two continuous sampling periods.
The initial state of the centers of the L group targets calculated in step 6 is used as a cushion for predicting the center positions of the L group targets in the next sampling period, i.e., the fifth time.
And 7, the radar data processing module starts to track the L group targets by taking the initial states of the centers of the L group targets obtained in the step 6 as a reference, the total tracking steps are set to be K, and K =0 is initialized.
In step 7, the sampling time for starting tracking of the L group targets is the fifth time, and L group targets at the next time are sequentially tracked.
Step 8, after receiving all echo signals in the next sampling period, the radar receiver executes the steps 1 to 4, so that equivalent central vectors Z of F group targets in the next sampling period are obtained c Then Z is c =(z 1 ,z 2 ,…z f ,…,z F )。
According to the embodiment of the invention, before radar filtering is carried out for each tracking, firstly, a radar receiver receives echo signals of all group targets in the next sampling period to be tracked, and then, the step 1 to the step 4 are carried out to obtain equivalent central vectors Z of all the group targets to be tracked c Preparation is made for the step 9 calculation.
Step 9, equivalent central vectors Z of F group targets in the next sampling period c And predicted central position vectors Z corresponding to the L group targets in the next sampling period c Performing correlation calculation to obtain a correlation matrix C, and obtaining predicted central position vectors Z of the L group targets in the next sampling period c ' calculated in substep 3 of this step 9;
the step 9 specifically includes the following three substeps:
substep 2, when F > L, indicating that separation occurs in L tracked group targets, then continuing to execute step 4 from step 5 to step 6 to recalculate initial states of centers of F group targets, and then executing substep 3 in step 9;
and a substep 3 of adding 1 to k and calculating the correlation matrix C.
According to the embodiment of the present invention, the relationship between F and L is first determined before calculating the correlation matrix, because the separation of group targets occurs during actual tracking: if F = L indicates that the tracked group targets are not separated, the following calculation can be continued; if F & gtL indicates that the tracked group targets are separated, the equivalent centers of the F group targets and the initial states of the F group targets need to be recalculated, and the subsequent continuous group target tracking is laid.
Preferably, the specific calculation process of the sub-step 3 is as follows:
9a) The value of k is judged, and if k =1, substeps 9 b) and 9 d) are performed in sequence, otherwise, substeps 9 c) and 9 d) are performed in sequence.
After the step 9 a) judges the value of k, two calculations are performed, and in practice, only when the group target starts to track, the substeps 9 b) and 9 d) are performed in sequence, because the filtering state without radar filtering tracking in the sampling period is tracked for the first time, and the substeps 9 c) and 9 d) are performed during the subsequent tracking, and the central position of the group target is predicted by using the filtering state with radar filtering tracking.
9b) And 6, predicting the central positions of the L group targets in the next sampling period by using the initial states of the centers of the L group targets in the step 6 as a reference and adopting the following formula, so as to obtain the predicted central positions of the L group targets in the next sampling period:
X_pre (l) =X_initial (l) +VX_initial (l) ·Δt
Y_pre (l) =Y_initial (l) +VY_initial (l) ·Δt
wherein, X _ pre (l) 、Y_pre (l) Respectively representing the X-direction coordinate and the Y-direction coordinate of the prediction center of the ith group target in the next sampling period, wherein delta t represents the time interval between two continuous sampling periods;
the predicted central position vector Z of the L group targets in the next sampling period can be obtained c ′,Z c ′=(z 1 ′,z 2 ′,…z l ′,…,z L ') wherein z is l ′=(X_pre (l) ,Y_pre (l) )。
9c) Predicting the center positions of the L group targets in the next sampling period by adopting the following formula according to the filtering states of the centers of the current L group targets, thereby obtaining the predicted center position of the ith group target in the next sampling period:
X_pre (l) =X_f (l) +VX_f (l) ·Δt
Y_pre (l) =Y_f (l) +VY_f (l) ·Δt
wherein, X _ pre (l) 、Y_pre (l) X-direction coordinates and Y-direction coordinates respectively representing the prediction center of the ith group target in the next sampling period, X _ f (l) 、Y_f (l) Respectively showing the coordinate in the X direction and the coordinate in the Y direction after the center of the current ith group target is filtered, and VX _ f (l) 、VY_f (l) Respectively representing the speed in the X direction and the speed in the Y direction after the center of the current ith group target is filtered, wherein delta t represents the sampling time interval of two continuous frames of observation information;
then the predicted central position vector Z of L group targets in the next sampling period can be obtained c ′,Z c ′=(z 1 ′,z 2 ′,…z l ′,…,z L ') wherein z is l ′=(X_pre (l) ,Y_pre (l) )。
9d) Equivalent central vector Z of F group targets in next sampling period c And predicted central position vectors Z corresponding to the L group targets in the next sampling period c ' performing correlation calculation:
setting the threshold value U of a rectangular wave gate x 、U y Determining the element C of the correlation matrix C according to fl :
Wherein, c fl Represents the elements of the ith row and ith column in the incidence matrix C, | | represents the absolute value, & represents the AND operation, xc f ,Yc f And respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the f-th group target in the rectangular coordinate system in the next sampling period.
And 11, judging the value of K, executing the steps 8 to 10 if K is less than K, and ending the tracking if K is less than K.
The effect of the method provided by the embodiment of the invention is verified through a simulation experiment as follows:
simulation experiment environment:
the experimental environment is as follows: inter (R) Core (TM) i7-7700CPU @3.60HGz, 64-bit Windows operating system and MATLAB 2014a simulation software.
Experimental data:
the total number Q =24 of the targets in one sampling period received by the radar receiver, and the total number L =3 of the group targets, each group comprises 8 targets, the number of the actual observed targets detected by the radar is 3-4, and the group targets 1,2 and 3 start from [ -8000-2000],[-80002000],[-6000-4000]Near, range error is σ r =20m, angle measurement error σ θ =0.1rad. Radar range resolution of lambda r =20, angular resolution λ θ =0.01rad。
The experimental results are as follows:
1. fig. 2 and fig. 3 respectively show schematic diagrams of a group target image before and after image form processing according to an embodiment of the present invention, that is, fig. 2 and fig. 3 respectively show a group target image before and after expansion processing of the group target image, and it can be known from the two images that the image pixel units of the same group target form a connected region after processing, so that the center of each connected region can be calculated and taken as the equivalent center of the corresponding group target, thereby avoiding the phenomenon that the center of the group target is shifted when a part of the observed image is missing, minimizing the influence of the missing image observation, ensuring the accuracy of the group target center calculation, and making the group target center tracking more stable.
2. As shown in fig. 4, which is an equivalent center trajectory of a tracking group target based on image morphology processing according to an embodiment of the present invention, and fig. 5, which is a real trajectory observed at the center of a group target according to an embodiment of the present invention, it can be known from the two figures that the equivalent center trajectory of the tracking group target based on image morphology processing matches with the real trajectory observed at the center of the group target very well, and no large deviation or tracking loss occurs, which indicates that the equivalent center of the group target obtained based on image morphology processing according to an embodiment of the present invention is reliable.
3. Fig. 6 and 7 are schematic diagrams of a distance error and an azimuth error of a center track of a group target 1 according to an embodiment of the present invention, fig. 8 and 9 are schematic diagrams of a distance error and an azimuth error of a center track of a group target 2 according to an embodiment of the present invention, respectively, and fig. 10 and 11 are schematic diagrams of a distance error and an azimuth error of a center track of a group target 3 according to an embodiment of the present invention, respectively, as can be seen from the diagrams, a distance error of a center of the group target 1 is-20 m to 25m, and an azimuth error is-0.3 degree to 0.5 degree; the distance error of the center of the group target 2 is-25 m-20 m, and the azimuth angle error is-0.5 degrees; the distance error of the center of the group target 3 is-20 m-40 m, and the azimuth error is-0.5 degree-0.3 degree, which shows that the embodiment of the invention adjusts the radar filtering tracking parameter through the incidence matrix between the group target equivalent center set and the group target prediction center position, so that the filtering tracking performance has good tracking effect in both distance and angle, and the good tracking effect can be still kept under the condition of partial observation missing of the group target.
As known from simulation experiments, when the method for tracking the group targets based on image morphological processing is used in the technical field of radar tracking, the method not only ensures the calculation accuracy of the group target center, but also can well track the group targets under the condition that the group targets are partially lost, and has valuable practicability.
Those of ordinary skill in the art will understand that: all or part of the steps of implementing the method embodiments may be implemented by hardware related to program instructions, and the program may be stored in a computer-readable storage medium, and when executed, executes the steps including the method embodiments; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (8)
1. A group target tracking method based on image morphological processing is characterized by comprising the following steps:
step 1, processing echo signals received by a radar receiver in a sampling period to obtain the total number Q of targets in all the echo signals of the radar in the sampling period, the distance and the azimuth angle measured by each target, and calculating the X-direction coordinate and the Y-direction coordinate of the distance and the azimuth angle measured by each target in a two-dimensional rectangular coordinate system through rectangular coordinate conversion;
step 2, establishing an initial image matrix I and calculating a binary image matrix I';
step 3, sequentially carrying out image morphological processing and marking processing on the binary image matrix I' to obtain the number L of group targets and an image matrix Ipm containing different marking numbers L, wherein L =1, 2., a, \8230;, L;
step 4, calculating equivalent central vectors Z, Z = (Z) of L group targets in the sampling period according to the image matrix Ipm containing different mark numbers L 1 ,z 2 ,…z l ,…,z L ) Wherein z is l =(Xc l Yc l ),Xc l ,Yc l Respectively representing the sampling periodThe equivalent center of the inner first group target is in the X-direction coordinate and the Y-direction coordinate in the rectangular coordinate system;
step 5, calculating a track set H = { H } of the L group targets by adopting a logical method inference according to equivalent center coordinates of the L group targets in a plurality of sampling periods 1 ,H 2 ,…,H l ,…,H L In which H l A stable track for the first group of targets;
step 6, calculating the initial states of the L group target centers, wherein the initial state of the ith group target center is as follows:
wherein, X _ st (l) Represents the initial state of the ith cluster target center, X _ initial (l) 、Y_initial (l) X-direction coordinates and Y-direction coordinates respectively representing the center of the first group target in the initial state, VX _ initial (l) 、VY_initial (l) Respectively representing the X-direction speed and the Y-direction speed of the center of the first group target in the initial state;
step 7, the radar data processing module starts to track the L group targets by taking the initial states of the centers of the L group targets obtained in the step 6 as a reference, the total tracking steps are set to be K, and K =0 is initialized;
step 8, after receiving all echo signals in the next sampling period, the radar receiver executes the steps 1 to 4, so that equivalent central vectors Z of F group targets in the next sampling period are obtained c Then Z is c =(z 1 ,z 2 ,…z f ,…,z F );
Step 9, equivalent central vectors Z of F group targets in the next sampling period c And predicted central position vectors Z corresponding to the L group targets in the next sampling period c Performing correlation calculation to obtain a correlation matrix C, and obtaining predicted central position vectors Z of the L group targets in the next sampling period c ' calculated in substep 3 of this step 9;
the step 9 specifically includes the following three substeps:
substep 1, judging the relation between F and L: if F > L, performing substep 2; if F = L, performing substep 3;
substep 2, when F > L, indicating that the tracked L group targets are separated, then continuing to execute step 4 from step 5 to step 6 to recalculate the initial states of the centers of the F group targets, and then executing substep 3 in step 9;
substep 3, adding 1 to k, and calculating the incidence matrix C;
step 10, obtaining the element C in the incidence matrix C according to the step 9 fl The corresponding group target filtering tracking track is adjusted by the numerical value: for c fl If the equivalent center of the F-th group target in the next sampling period is associated with the prediction center of the l-th group target, applying the equivalent center coordinate of the F-th group target to the track of the group target in the next sampling period for filtering and tracking to obtain the filtering tracks of the current F group targets;
meanwhile, taking the filter tracks of the current F group targets as the filter states for predicting the center positions of the F group targets in a new next sampling period on one hand, and as the tracking results of the F group targets on the other hand;
and 11, judging the value of K, executing the steps 8 to 10 if K is less than K, and ending the tracking if K is less than K.
2. The method for tracking the group target based on the image morphological processing as claimed in claim 1, wherein the step 1 comprises the following sub-steps:
the substep 1, sequentially carrying out pulse pressure, constant false alarm detection and monopulse angle measurement processing on all echo signals received by a radar receiver in a sampling period, thereby obtaining the total number Q of targets, the distance measured by each target and the azimuth angle measured by each target in all the echo signals received by the radar in the sampling period;
and substep 2, converting the distance and the azimuth angle measured by each target into an X-direction coordinate and a Y-direction coordinate of each target in a two-dimensional rectangular coordinate system by adopting a rectangular coordinate conversion method according to the following formula:
x d =R d cosθ d
y d =R d sinθ d
wherein x is d 、y d Respectively representing the X-direction coordinate and the Y-direction coordinate of the measured distance and the measured azimuth angle of the d-th target in a two-dimensional rectangular coordinate system, R d Represents the distance, θ, of the d-th target measurement d Indicating the azimuth angle of the d-th target measurement.
3. The method for tracking the group target based on the image morphological processing as claimed in claim 1, wherein the step 2 comprises the following sub-steps:
substep 1, setting the range of the target monitoring area as [ X min, X max ] × [ Y min, Y max ], wherein X min and X max respectively represent the minimum value of the target monitoring area range in the X direction and the maximum value of the target monitoring area range in the X direction, and Y min and Y max respectively represent the minimum value of the target monitoring area range in the Y direction and the maximum value of the target monitoring area range in the Y direction;
meanwhile, the size of a pixel unit of an image is set to be x _ im multiplied by y _ im, wherein x _ im and y _ im respectively represent the width and the height of the pixel unit of the image;
and a substep 2 of calculating the dimensions of the rows and columns of the image matrix I, wherein the calculation formula is as follows:
wherein, P represents the dimension of the row of the image matrix I, and N represents the dimension of the column of the image matrix I;
substep 3, initializing all pixel units in the image matrix I, wherein the initial value is 0;
and substep 4, calculating the X-direction coordinate and the Y-direction coordinate of all the target measurements in the step 1, which respectively correspond to the X-direction coordinate and the Y-direction coordinate in the image matrix I, wherein the calculation formula is as follows:
wherein m =1,2, \ 8230;, Q, X m _Im、Y m Im respectively represents the X-direction coordinate and the Y-direction coordinate of the mth target measurement in the image matrix I, and X is m _dkr、Y m Anddkr respectively represents the X-direction coordinate and the Y-direction coordinate of the m-th target measurement,<>the method comprises the following steps of (1) indicating a downward integer, wherein X min represents the minimum value of a target monitoring area range in the X direction, and Y min represents the minimum value of the target monitoring area range in the Y direction;
and a substep 5, finding a coordinate position corresponding to each target measurement in the image matrix I according to the X-direction coordinate and the Y-direction coordinate of all the obtained target measurements in the image matrix I, and updating and assigning the initial value 0 of the corresponding pixel unit at the coordinate position to be 1, so as to obtain a new image matrix I 'containing all the target measurements, wherein the image matrix I' is a binary image matrix.
4. The method for tracking the group target based on the image morphological processing as claimed in claim 1, wherein the step 3 specifically comprises:
firstly, the binary image matrix I' is morphologically processed:
setting the structural element B as a matrix of a disk-shaped structure with the radius of 8 according to an image expansion algorithmPerforming expansion operation on the binary image matrix I' to obtain a binary image after edge expansionA matrix Ip and a plurality of connected regions are formed in the binary image matrix Ip; the dimensions of the rows and the columns in the binary image matrix Ip are respectively consistent with the dimensions of the rows and the columns in the image matrix I';
secondly, a connected domain marking method is adopted to perform marking calculation on a plurality of connected regions in the binary image matrix Ip, and the marking calculation process comprises the following steps:
4a) Scanning the binary image matrix Ip line by line in a line sequence, and if one or more pixel units in all pixel units in the first line are 1, executing 4 b) to 4 d); if all the pixel units in the first row are 0, continuing to scan the second row, and if all the pixel units in the second row are 0, continuing to scan the third row until the pixel unit in the scanned row appears 1 for the first time, and executing 4 b) to 4 d);
4b) Scanning a binary image matrix Ip line by line, forming a sequence with continuous pixel units being 1 in each line, and recording a starting point, an end point and a line number of each sequence;
for a row with a pixel unit of 1 appearing for the first time in all rows of a binary image matrix Ip, sequentially marking all sequences in the row, wherein the corresponding marking number l is l =1,2, \8230, a;
4c) Marking all sequences in the next row in sequence on the basis of the mark numbers of all sequence marks in the row of 4 b), wherein if the s-th sequence has a superposition area with one sequence in the row, the mark number of the s-th sequence mark is the same as the mark number of the sequence mark in the row; if the s-th sequence has an overlapping area with two or more sequences in the row, the mark number of the s-th sequence mark is the same as the smallest mark number of the mark numbers of the two or more sequence marks in the row, and the mark numbers of the two or more sequence marks in the row are recorded as an equivalent pair; if the s-th sequence has no overlapping area with all the sequences in the row, the mark number marked by the s-th sequence is a new mark number, when s =1, the new mark number l = a +1, and when s ≠ 1, the new mark number is sorted according to the new mark number marked by the next row of sequences;
by analogy, all sequences in all the following rows in the binary image matrix Ip are marked in sequence;
4d) And (3) respectively equating the mark numbers of the different equivalent pairs recorded in the step 4 c) to the mark number of the minimum value in the mark numbers, finishing the calculation of eliminating the equivalent pairs, traversing all the previous marks and re-marking all the connected regions in a natural number sequence to obtain an image matrix Ipm containing different mark numbers L, wherein the maximum value of the mark is the number L of the connected regions, the number of the connected regions is the number of the group targets, and the dimensions of the rows and the columns of the image matrix Ipm are respectively consistent with the dimensions of the rows and the columns in the image matrix Ip.
5. The method for tracking the group target based on the image morphological processing as claimed in claim 1, wherein the step 4 comprises the following sub-steps:
substep 1, calculating the coordinates of the centers of all connected regions in the image matrix Ipm according to the following formula:
wherein, the first and the second end of the pipe are connected with each other,the X-direction coordinate representing the center of the l-th connected component in the image matrix Ipm,y-direction coordinate, x, representing the center of the l-th connected region in the image matrix Ipm 1n ,y lj Respectively representing the X-direction coordinate and the Y-direction coordinate of the nth pixel unit in the ith communication region, and lr and lc respectively representing the X-direction pixel unit marked as l in the pixel unit in the image matrix IpmTotal number of Y-direction pixel units, S l Is the total number of pixel units in the first connected region;
substep 2, calculating the coordinates of the equivalent centers of the L group targets according to the following formula:
wherein Xc l ,Yc l Respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the first group target in a rectangular coordinate system;
substep 3, obtaining equivalent center vectors Z = (Z is Z) of the L group targets in the sampling period according to the coordinates of the equivalent centers of the L group targets 1 ,z 2 ,…z l ,…,z L ) Wherein z is l =(Xc l ,Yc l )。
6. The method for tracking the group target based on the image morphological processing as claimed in claim 1, wherein the specific process of the step 5 is as follows:
if one sampling period is one sampling time, the multiple sampling periods are multiple sampling times, and the radar sampling time is the first T times, then when T =1, T =2, T =3, and T =4, steps 1 to 4 are respectively executed, and the equivalent center vectors of L group targets at the corresponding time are obtained: let T = T, T =1,2,3,4, then the equivalent center vector of L group targets at the T-th time is Z t ,Z t =(z (t,1) ,z (t,2) ,…,z (t,L) ) Wherein z is (t,l) =(Xc t,l ,Yc t,l ),Xc t,l ,Yc t,l Respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the ith group target in a rectangular coordinate system at the tth moment;
therefore, the equivalent center coordinates of the L group targets at the first four moments are utilized, and the stable tracks of the L group targets are inferred by adopting a logic method:
firstly, establishing an initial relevant wave gate by using the speed method for the equivalent centers of L group targets in a first moment, and establishing a possible track set for the equivalent centers of the L group targets in a second moment falling into the initial relevant wave gate;
secondly, extrapolating each subset track in the possible track set, taking an extrapolation point as a center, and determining the size of the wave gate at the moment by the covariance of track extrapolation errors: if the equivalent center of the group target falling into the correlation wave gate exists in the third moment, storing the equivalent center of the group target with the nearest extrapolation point into the corresponding possible track subset; if the equivalent center of no group target falls into the relevant wave gate, deleting the possible subset track from the possible track set;
finally, continuously extrapolating the remaining possible flight paths, if a group equivalent center falling into a relevant wave gate exists at the fourth moment, storing the group equivalent center closest to the extrapolated point into the corresponding possible flight path set, and determining the flight path as a stable flight path; if no group equivalent center falls into the relevant wave gate, deleting the possible track from the possible track set;
after the processing according to the logic method, a stable track set H of L group targets can be obtained, wherein H = { H = 1 ,H 2 ,…,H l ,…,H L In which H l Is the stable track of the ith group of targets and is also the subset formed by the equivalent center coordinate positions of the ith group of targets at the first four moments, then H l Is expressed as H l ={h (l,1) ,…,h (l,t) ,…h (l,4) In which the element h (l,t) Is the equivalent center coordinate position of the ith group target at the t-th time, i.e.Xc (l,t) 、Yc (l,t) Respectively representing the X power of the equivalent center of the first group target at the t time in a rectangular coordinate systemA directional coordinate and a Y-direction coordinate.
7. The method for tracking the group target based on the image morphological processing as claimed in claim 1, wherein the step 6 specifically comprises:
the initial state of the ith cluster target center is calculated according to the following formula:
wherein Xc (l,T) 、Yc (l,T) Respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the ith group target at the Tth time in a rectangular coordinate system, xc (l,t+1) 、Yc (l,t+1) Respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the ith group target at the t +1 th moment in a rectangular coordinate system, and delta t represents the time interval between two continuous sampling periods.
8. The method for tracking group targets based on image morphological processing as claimed in claim 1, wherein the sub-step 3 of the step 9 specifically comprises:
8a) Judging the value of k, and if k =1, sequentially performing substeps 8 b) and 8 d), otherwise, sequentially performing substeps 8 c) and 8 d);
8b) And 6, predicting the central positions of the L group targets in the next sampling period by using the initial states of the centers of the L group targets in the step 6 as a reference and adopting the following formula, so as to obtain the predicted central position of the ith group target in the next sampling period:
X_pre (l) =X_initial (l) +VX_initial (l) ·Δt
Y_pre (l) =Y_initial (l) +VY_initial (l) ·Δt
wherein, X _ pre (l) 、Y_pre (l) Respectively representing the X-direction coordinate and the Y-direction coordinate of the prediction center of the ith group target in the next sampling period, and delta t representing the time between two consecutive sampling periodsSpacing;
then the predicted central position vector Z of L group targets in the next sampling period can be obtained c ′,Z c ′=(z 1 ′,z 2 ′,…z l ′,…,z L ') wherein z is l ′=(x_pre (l) ,Y_pre (l) );
8c) Predicting the center positions of the L group targets in the next sampling period by adopting the following formula according to the filtering states of the centers of the current L group targets, thereby obtaining the predicted center position of the ith group target in the next sampling period:
X_pre (l) =X_f (l) +VX_f (l) ·Δt
Y_pre (l) =Y_f (l) +VY_f (l) ·Δt
wherein, X _ pre (l) 、Y_pre (l) X-direction coordinates and Y-direction coordinates respectively representing the prediction center of the ith group target in the next sampling period, X _ f (l) 、Y_f (l) Respectively represent the X-direction coordinate and the Y-direction coordinate after the current l-th group target center is filtered, VX _ f (l) 、VY_f (l) Respectively representing the speed in the X direction and the speed in the Y direction after the center of the current ith group target is filtered, wherein delta t represents the sampling time interval of two continuous frames of observation information;
the predicted central position vector Z of the L group targets in the next sampling period can be obtained c ′,Z c ′=(z 1 ′,z 2 ′,…z l ′,…,z L ') wherein z is l ′=(X_pre (l) ,Y_pre (l) );
8d) Equivalent central vector Z of F group targets in next sampling period c And the predicted central position vector Z corresponding to the L group targets in the next sampling period c ' performing correlation calculation:
setting a rectangular wave gate threshold value U x 、U y Determining the element C of the correlation matrix C according to fl :
Wherein, c fl Represents the element in the ith row and ith column in the correlation matrix C, | | | represents taking the absolute value,&represents and operation, xc f ,Yc f And respectively representing the X-direction coordinate and the Y-direction coordinate of the equivalent center of the f-th group target in the rectangular coordinate system in the next sampling period.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910057384.3A CN109814074B (en) | 2019-01-22 | 2019-01-22 | Group target tracking method based on image morphological processing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910057384.3A CN109814074B (en) | 2019-01-22 | 2019-01-22 | Group target tracking method based on image morphological processing |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109814074A CN109814074A (en) | 2019-05-28 |
CN109814074B true CN109814074B (en) | 2022-11-18 |
Family
ID=66603610
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910057384.3A Active CN109814074B (en) | 2019-01-22 | 2019-01-22 | Group target tracking method based on image morphological processing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109814074B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111208501B (en) * | 2019-12-30 | 2022-06-03 | 广州市番禺奥莱照明电器有限公司 | Multi-target aggregation method, device, equipment and storage medium based on radar detection |
CN113030943B (en) * | 2021-03-05 | 2023-08-18 | 中国人民解放军空军工程大学航空机务士官学校 | Multi-target tracking algorithm based on monopulse radar signal acquisition azimuth range profile |
CN113658220B (en) * | 2021-07-23 | 2024-01-19 | 哈尔滨工业大学 | Group target tracking method based on fusion space clustering segmentation and multiple cost matrix association |
CN114859300B (en) * | 2022-07-07 | 2022-10-04 | 中国人民解放军国防科技大学 | Radar radiation source data stream processing method based on graph connectivity |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2004233136A (en) * | 2003-01-29 | 2004-08-19 | Mitsubishi Electric Corp | Target tracking device |
CN108303692A (en) * | 2018-01-30 | 2018-07-20 | 哈尔滨工业大学 | A kind of multi-object tracking method of solution doppler ambiguity |
CN109031279A (en) * | 2018-06-23 | 2018-12-18 | 中国人民解放军海军航空大学青岛校区 | Multiple multiple targets trackings with shape information |
-
2019
- 2019-01-22 CN CN201910057384.3A patent/CN109814074B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2004233136A (en) * | 2003-01-29 | 2004-08-19 | Mitsubishi Electric Corp | Target tracking device |
CN108303692A (en) * | 2018-01-30 | 2018-07-20 | 哈尔滨工业大学 | A kind of multi-object tracking method of solution doppler ambiguity |
CN109031279A (en) * | 2018-06-23 | 2018-12-18 | 中国人民解放军海军航空大学青岛校区 | Multiple multiple targets trackings with shape information |
Also Published As
Publication number | Publication date |
---|---|
CN109814074A (en) | 2019-05-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109814074B (en) | Group target tracking method based on image morphological processing | |
EP2062067B1 (en) | Method of and device for tracking an object | |
WO2012009947A1 (en) | Device and method of signature-driven multi-target tracking | |
CN107861123A (en) | A kind of through-wall radar is under complex environment to the method for multiple mobile object real-time tracking | |
CN114049382B (en) | Target fusion tracking method, system and medium in intelligent network connection environment | |
CN108957437B (en) | Radar target tracking method based on environmental situation analysis | |
CN107436434B (en) | Track starting method based on bidirectional Doppler estimation | |
CN115840221B (en) | Method for achieving target feature extraction and multi-target tracking based on 4D millimeter wave radar | |
Cao et al. | Correlation-based tracking of multiple targets with hierarchical layered structure | |
CN111929653B (en) | Target detection and tracking method and system based on unmanned ship navigation radar | |
CN114063056A (en) | Ship track fusion method, system, medium and equipment | |
CN111259332B (en) | Fuzzy data association method and multi-target tracking method in clutter environment | |
CN114779204B (en) | Radar-based target amplitude least square tracking processing method and system | |
CN113702979A (en) | Cross-region target tracking trajectory segment space-time splicing method | |
CN115220002B (en) | Multi-target data association tracking method and related device for fixed single station | |
CN107705319B (en) | Aerospace background-based small target real-time detection tracking method | |
CN113721237B (en) | Multi-membership-degree target intelligent matching algorithm | |
Kim et al. | Dynamic model‐based feature aided data association filter in target tracking | |
Zhang et al. | Learning to multi-target tracking in dense clutter environment with JPDA-recurrent neural networks | |
CN112748735B (en) | Extended target tracking method introducing color features | |
CN115100565A (en) | Multi-target tracking method based on spatial correlation and optical flow registration | |
CN113484866A (en) | Multi-target detection tracking method based on passive sonar azimuth history map | |
Wang et al. | Research on multi-maneuvering target tracking JPDA algorithm | |
CN112379366A (en) | Radar target tracking algorithm based on particle swarm filtering | |
CN102707278B (en) | Multi-target tracking method for singular value decomposition |
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 |