CN106443618B - A kind of high resolution radar Data Association based on standard deviation ellipse parameter auxiliary - Google Patents
A kind of high resolution radar Data Association based on standard deviation ellipse parameter auxiliary Download PDFInfo
- Publication number
- CN106443618B CN106443618B CN201610811607.7A CN201610811607A CN106443618B CN 106443618 B CN106443618 B CN 106443618B CN 201610811607 A CN201610811607 A CN 201610811607A CN 106443618 B CN106443618 B CN 106443618B
- Authority
- CN
- China
- Prior art keywords
- target
- standard deviation
- deviation ellipse
- subclass
- similarity
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/415—Identification of targets based on measurements of movement associated with the target
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/50—Systems of measurement based on relative movement of target
- G01S13/58—Velocity or trajectory determination systems; Sense-of-movement determination systems
Abstract
The invention discloses a kind of high resolution radar Data Associations based on standard deviation ellipse parameter auxiliary.The present invention is the plot-track Association Algorithm based on standard deviation ellipse, and more other feature auxiliary association algorithms have preferably association accuracy.The invention firstly uses test point mark group of the standard deviation ellipse to target to carry out clustering processing, and different Targets Dots clustering classes is the ellipse of different characteristic, proposes the association process of the concept auxiliary interframe target of elliptical similarity later.
Description
Technical field
The invention belongs to high resolution radar technical fields, and in particular to a kind of high score based on standard deviation ellipse parameter auxiliary
Distinguish radar track correlating method.
Background technique
Wideband High Resolution Radar is the important development direction of modern radar, and high resolution radar also provides more target letters
Breath assists association process using clarification of objective information, can greatly improve correlation accuracy.Herein, high-resolution radar refers to
The minimum actual range for two objects that can be differentiated is in Centimeter Level and radar below.Traditional Narrow-band Radar carries out target spy
When survey, target size is usually less than radar resolution, therefore target is usually modeled as a point target, is based on point target model
Propose arest neighbors, probability data interconnection, Joint Probabilistic Data Association scheduling algorithm, wherein nearest neighbor algorithm be most simply be also answer
With a kind of most wide data association algorithm.Compared to traditional Narrow-band Radar, the testing result of Wideband High Resolution Radar target is not
It can be represented with simple point model, so that arest neighbors association algorithm is restricted.Present feature assists association algorithm master
It concentrates on being associated with using auxiliary such as the one-dimensional range profile feature of target, amplitude characteristic, Doppler Feature, polarization characteristics, target
These be generally characterized by any time, the factors variation such as target state, and have certain fluctuating, it is accurate to influence association
Degree.In view of target space structure not at any time, motion state variation, and do not rise and fall, therefore the present invention proposes a kind of base
In the auxiliary association algorithm of the spatial structure characteristic of target, the distribution characteristics of target detection point is identified using standard deviation ellipse,
And then assist track association.
Summary of the invention
In view of this, the present invention provides a kind of high resolution radar track association sides based on standard deviation ellipse parameter auxiliary
Method can more accurately carry out track association in complex situations.
High resolution radar shows as the point mark group being distributed in a certain range to the testing result of target, puts the distribution shape of mark group
Shape is related with the space structure of target.The invention firstly uses test point mark group of the standard deviation ellipse to target to carry out at cluster
Reason, different Targets Dots clustering classes are the ellipse of different characteristic.The concept auxiliary interframe of elliptical similarity is proposed later
The association process of target.
It is including following the present invention relates to a kind of high resolution radar Data Association based on standard deviation ellipse parameter auxiliary
Step:
Step 1, target and its corresponding standard deviation ellipse for obtaining present frame and next frame, specifically include following sub-step
It is rapid:
Step 1.1 establishes standard deviation ellipse: on the range Doppler plan view of present frame and next frame, for comprising
The set of all high resolution radar test points, establishes standard deviation ellipse.
Target number in step 1.2, judgement set: it calculates and crosses the center of circle in standard deviation ellipse and be parallel to distance samples list
The length l of the line segment mn of first axismn, and it is compared with target in the maximum distribution length Dr of radar radially；If Dr
≤lmn, then judge to detect in point set comprising multiple targets, execute step 1.3；Otherwise, judge to detect in point set only comprising one
A target executes step 1.4.
Step 1.3, subclass divide: the test point that abscissa is less than or equal to center of circle abscissa is divided into a subset
It closes, the test point that abscissa is greater than center of circle abscissa is divided into another subclass；Then it is directed to each subclass, using step
Rapid 1.1~1.3 method is established standard deviation ellipse model, the judgement of target number and subclass and is divided, respectively until each subset
Closing all only includes a target, executes step 1.4.
Step 1.4, aggregation test point: including the set or subclass of 1 target for each, to the set or subset
Conjunction is assembled, and the point for choosing amplitude maximum in set or subclass is pressed as target or by the test point in set or subclass
Amplitude is weighted and averaged as target, thus to obtain the target and its corresponding standard deviation ellipse of present frame and next frame.
Step 2, the oval feature parameter matrix and measurement matrix for establishing target: ellipse for k-th of target of present frame
Circle characteristic parameter Tk(i) are as follows:
Tk(i)=[σk1(i) σk2(i) θk(i) nk(i)]T (1)
Wherein, T represents transposition, σk1(i)、σk2(i)、θk(i) and nkIt (i) is respectively the corresponding standard deviation ellipse of the target
Test point number in long axis, short axle, deflection and ellipse；
Measurement matrix Z of the target in the i-th frame of present framek(i) are as follows:
Zk(i)=[xk(i) yk(i) zk(i) Vkx(i) Vky(i) Vkz(i) σk1(i) σk2(i) θk(i) nk(i)]T (2)
Wherein, xk(i)、yk(i) and zkIt (i) is respectively point of the target position under rectangular coordinate system in space on x, y, z axis
Amount；Vkx(i)、Vky(i) and VkzIt (i) is respectively component of the target velocity in x, y, z axis direction.
Step 3, screening candidate point mark: calculating measurement predictor according to the measurement matrix of present frame target k, to measure prediction
Elliptical wave door is established centered on value；From the target of the next frame obtained in step 1, screening falls into the mesh of the next frame of Bo Mennei
It is designated as the candidate point mark of target k.
Step 4 judges candidate point mark number: if not having candidate point mark, the failure of this track association terminates this pass
Connection；If only one candidate point mark, track directly is updated with the mark；If there is multiple candidate point marks, then follow the steps
5。
Step 5, calculate present frame target k and each candidate point mark between similarity:
Calculate separately the angle of the standard deviation ellipse of present frame target k and the standard deviation ellipse of each candidate point mark, long axis,
The similarity of short axle and test point number, with the corresponding candidate point mark of the maximum of the product of above-mentioned similarity more fresh target k's
Track.
Preferably, steps are as follows for the calculating of target similarity in the step 5:
Assuming that the characteristic parameter of the standard deviation ellipse of the standard deviation ellipse and candidate point mark m of present frame target k is respectively
Step 5.1: calculating angle similarity
Sθ=| cos (θk(i)-θm(i+1))| (4)
Step 5.2: calculating long axis similarity
Step 5.3: calculating short axle similarity
Step 5.4: calculating test point number similarity
Step 5.5: calculating target similarity
Step 5.6: choosing the corresponding candidate point mark of target similarity the maximum and update track.
The utility model has the advantages that
The spatial structure characteristic of target is more stable compared with other features, therefore the plot-track Association Algorithm based on standard deviation ellipse
More other feature auxiliary association algorithms have preferably association accuracy.
Detailed description of the invention
Fig. 1 is that the test point based on standard deviation ellipse assembles process.
Fig. 2 is the high resolution radar Data Association flow chart assisted based on standard deviation ellipse parameter.
Fig. 3 is to calculate target similarity flow chart.
Fig. 4 is the high resolution radar track association schematic diagram assisted based on standard deviation ellipse parameter.
Fig. 5 is the high resolution radar track association result assisted based on standard deviation ellipse parameter.
Specific embodiment
The present invention will now be described in detail with reference to the accompanying drawings and examples.
The present invention is primarily based on standard deviation ellipse, and the test point mark for Doppler's plane of adjusting the distance is assembled, and calculates each
A elliptical characteristic parameter；The track association that adjacent interframe tracks target is carried out by calculating oval similarity later.
1, test point is assembled based on standard deviation ellipse
Test point aggregation process based on standard deviation ellipse is as shown in Fig. 1:
Step 1, target and its corresponding standard deviation ellipse for obtaining present frame and next frame, specifically includes the following steps:
Step 1.1, the standard deviation ellipse for establishing test point
Assuming that high resolution radar target is after tested in range Doppler plane (with distance samples unit for horizontal seat
Mark, Doppler's channel signal be the plane of the composition of ordinate) forming quantity be n test point, these detect point sets be
G={ p1(r1,d1),p2(r2,d2),...,pj(rj,dj),..,pn(rn,dn)},1≤j≤n (1)
Wherein, riAnd diRespectively test point piThe distance samples unit and Doppler's channel number at place.
Standard deviation ellipse is established to the test point in set, calculated the length for being parallel to r axis segment mn in the oval center of circle
lmn, r axis is the horizontal axis of range Doppler plane.
Step 1.2: judgement detection point set includes target number
Because distance unit number shared by single target is limited, it is possible to pass through the length l of line segment mnmnTo judge current collection
Whether interior test point belongs to the same target, realizes the aggregation of test point, detailed process:
SentenceIt is disconnectedlmnWhether following condition is met:
lmn≤Dr (2)
Wherein, Dr is maximum distribution length of the target in radar radially；If conditions are not met, then determine include in set G
The test point of multiple targets executes step 1.3；Otherwise determine to execute step only comprising the test point of a target in set G
1.4。
Step 1.3: dividing detection point set is two subclass
If including the test point of multiple targets in set G, pass through following Rule of judgment
rj≤C1 (3)
Set G is divided into two subclass, wherein C1For projection of the oval center of circle on r axis；Then it is directed to every height
Set repeats the standard deviation ellipse that step 1.1~1.3 calculate two subclass, until the standard that the test point of each set is established
Poor elliptic parameter lmnAll meet formula (2).
Step 1.4: aggregation test point
Under conditions of meeting formula (2), that is, the test point in the ellipse is thought from a target, selection standard difference is oval
The point of interior amplitude maximum is weighted and averaged by amplitude as target as target or by the test point in oval circle.
Take the corresponding elliptical long axis σ of k-th of target of present framek1(i), short axle σk2(i), deflection θk(i) it and detects
Point number nk(i) it is used as oval feature parameter Tk(i),
Tk(i)=[σk1(i) σk2(i) θk(i) nk(i)]T (4)
It takes out the status information obtained after processing through radar return and the elliptical elliptic parameter establishes the target i-th
Frame measurement matrix Zk(i)。
Zk(i)=[xk(i) yk(i) zk(i) Vkx(i) Vky(i) Vkz(i) σk1(i) σk2(i) θk(i) nk(i)]T (5)
Wherein, T represents transposition, xk(i)、yk(i) and zkIt (i) is respectively x, y, z axis under the rectangular coordinate system in space of target position
On component.Vkx(i)、Vky(i) and VkzIt (i) is respectively component of the target velocity in x, y, z axis direction.
2, the high resolution radar Data Association based on standard deviation ellipse parameter auxiliary
Track association process based on standard deviation ellipse parameter auxiliary is as shown in Fig. 2:
For the either objective k obtained after test point aggregation, the track last state of the i-th frame tracking target is obtainedI.e.
White triangles shape in Fig. 4:
Zk(i)=[xk(i) yk(i) zk(i) Vkx(i) Vky(i) Vkz(i) σk1(i) σk2(i) θk(i) nk(i)]T
Step 2: calculating the measurement predictor of trackThat is the black dot in Fig. 4:
Wherein, H is measurement matrix, and F is dbjective state transfer matrix, and T is radar scanning interval.
Step 3: with measurement predictorCentered on establish elliptical wave door, i.e. oval dotted line in Fig. 4, from
In the target of the next frame obtained in step 1, the target that screening falls into the next frame of Bo Mennei is candidate point mark.Assuming that i+1
K-th of target measurement matrix of frame be
Zk(i+1)=[xk(i+1) yk(i+1) zk(i+1) Vkx(i+1) Vky(i+1) Vkz(i+1)
σk1(i+1) σk2(i+1) θk(i+1) nk(i+1)]T (9)
If Zk(i+1) meet:
vk T(i+1)Γ-1(i+1)vk(i+1)≤γ (10)
Then determine Zk(i+1) it is candidate point mark, is screened as shown in figure 4, sharing 3 targets as candidate point mark.
Wherein, γ is elliptical wave door size, can be by χ2Table distribution obtains, vk(i+1) it is new breath:
And:
For the measurement error of x, y, z, which is system registration test
Obtained in known quantity.
Step 4: if the target of next frame is not fallen into elliptical wave door, the failure of this track association then terminates this
Association；If only one target drops into elliptical wave door, track directly is updated with the measuring value；If there is multiple targets
It falls into elliptical wave door, then the similarity between present frame target and each candidate point mark is calculated, as shown in figure 4, similarity operator
Method flow chart is as shown in Fig. 3.
Steps are as follows for target similarity calculation:
The characteristic parameter of the standard deviation ellipse of the standard deviation ellipse and candidate point mark m of present frame target is respectively
Step 4.1: calculating angle similarity
Sθ=| cos (θk(i)-θm(i+1))| (14)
Step 4.2: calculating long axis similarity
Step 4.3: calculating short axle similarity
Step 4.4: calculating points similarity
Step 4.5: calculating target similarity
Wherein, SkIt is the target similarity of present frame target Yu candidate point mark m.Then step 4.6 is executed.
Step 4.6: choosing the maximum candidate point mark of target similarity and update track.
As shown in figure 4, the similarity between 3 candidate point marks and target, respectively S1、S2、S3.In S1、S2、S3It selects most
Big value, it is assumed that S3For maximum value, then S is used3Corresponding candidate point mark (i.e. gray triangles in Fig. 5) updates track.It is updated
Targetpath is as shown in Figure 5.
Claims (1)
1. a kind of high resolution radar Data Association based on standard deviation ellipse parameter auxiliary, which is characterized in that including following
Step:
Step 1, target and its corresponding standard deviation ellipse for obtaining present frame and next frame, specifically include following sub-step:
Step 1.1 establishes standard deviation ellipse: on the range Doppler plan view of present frame and next frame, for comprising all
The set of high resolution radar test point, establishes standard deviation ellipse；
Target number in step 1.2, judgement set: it calculates and crosses the center of circle in standard deviation ellipse and be parallel to distance samples unit shaft
Line segment mn length lmn, and it is compared with target in the maximum distribution length Dr of radar radially；If Dr≤
lmn, then judge to detect in point set comprising multiple targets, execute step 1.3；Otherwise, judge to detect in point set only comprising one
Target executes step 1.4；
Step 1.3, subclass divide: the test point that abscissa is less than or equal to center of circle abscissa be divided into a subset and is closed,
The test point that abscissa is greater than center of circle abscissa is divided into another subclass；Then it is directed to each subclass, using step
1.1~1.3 method is established standard deviation ellipse model, the judgement of target number and subclass and is divided, respectively until each subclass
All only include a target, executes step 1.4；
Step 1.4, aggregation test point: including the set or subclass of 1 target for each, to the set or subclass into
Row aggregation, the point for choosing amplitude maximum in set or subclass press amplitude as target or by the test point in set or subclass
It is weighted and averaged as target, thus to obtain the target and its corresponding standard deviation ellipse of present frame and next frame；
Step 2, the oval feature parameter matrix and measurement matrix for establishing target: oval special for k-th of target of present frame
Levy parameter Tk(i) are as follows:
Tk(i)=[σk1(i) σk2(i) θk(i) nk(i)]T (1)
Wherein, T represents transposition, σk1(i)、σk2(i)、θk(i) and nk(i) be respectively the corresponding standard deviation ellipse of the target length
Test point number in axis, short axle, deflection and ellipse；
Measurement matrix Z of the target in current i-th framek(i) are as follows:
Zk(i)=[xk(i) yk(i) zk(i) Vkx(i) Vky(i) Vkz(i) σk1(i) σk2(i) θk(i) nk(i)]T (2)
Wherein, xk(i)、yk(i) and zkIt (i) is respectively component of the target position under rectangular coordinate system in space on x, y, z axis；Vkx
(i)、Vky(i) and VkzIt (i) is respectively component of the target velocity in x, y, z axis direction；
Step 3, screening candidate point mark: measurement predictor is calculated according to the measurement matrix of present frame target k, is with measurement predictor
Elliptical wave door is established at center；From the target of the next frame obtained in step 1, the target for screening the next frame for falling into Bo Mennei is
The candidate point mark of target k；
Step 4 judges candidate point mark number: if not having candidate point mark, the failure of this track association terminates this secondary association；
If only one candidate point mark, track directly is updated with the mark；If there is multiple candidate point marks, 5 are thened follow the steps；
Step 5, calculate present frame target k and each candidate point mark between similarity:
Assuming that the characteristic parameter of the standard deviation ellipse of the standard deviation ellipse and candidate point mark m of present frame target k is respectively
Step 5.1: calculating angle similarity
Sθ=| cos (θk(i)-θm(i+1))| (4)
Step 5.2: calculating long axis similarity
Step 5.3: calculating short axle similarity
Step 5.4: calculating test point number similarity
Step 5.5: calculating target similarity
Step 5.6: choosing the corresponding candidate point mark of target similarity the maximum and update track.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610811607.7A CN106443618B (en) | 2016-09-08 | 2016-09-08 | A kind of high resolution radar Data Association based on standard deviation ellipse parameter auxiliary |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610811607.7A CN106443618B (en) | 2016-09-08 | 2016-09-08 | A kind of high resolution radar Data Association based on standard deviation ellipse parameter auxiliary |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106443618A CN106443618A (en) | 2017-02-22 |
| CN106443618B true CN106443618B (en) | 2019-01-04 |
Family
ID=58164436
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610811607.7A Active CN106443618B (en) | 2016-09-08 | 2016-09-08 | A kind of high resolution radar Data Association based on standard deviation ellipse parameter auxiliary |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106443618B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109711447A (en) * | 2018-12-19 | 2019-05-03 | 武大吉奥信息技术有限公司 | A kind of special population event early warning and monitoring method and device |
| CN111060906A (en) * | 2019-11-15 | 2020-04-24 | 驭势（上海）汽车科技有限公司 | Sensing data processing method and device, vehicle-mounted equipment and storage medium |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4558600B2 (en) * | 2005-07-20 | 2010-10-06 | 三菱電機株式会社 | Wake correlation integration device |
| CN105487061B (en) * | 2015-12-01 | 2018-04-17 | 中国人民解放军海军航空工程学院 | The multicharacteristic information fusion method of target data association |
| CN105510896B (en) * | 2015-12-03 | 2019-07-02 | 成都能通科技有限公司 | A kind of weighted nearest neighbor numeric field data correlating method of centralization multi-radar data processing |
| CN105913080B (en) * | 2016-04-08 | 2019-02-22 | 西安电子科技大学昆山创新研究院 | Joint tracking and classification method based on the motor-driven non-elliptical extension target of random matrix |
-
2016
- 2016-09-08 CN CN201610811607.7A patent/CN106443618B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN106443618A (en) | 2017-02-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106443598B (en) | Radar fence based on convolutional neural networks cooperates with track spoofing distinguishing disturbance method | |
| CN104183127B (en) | Traffic surveillance video detection method and device | |
| CN106125087A (en) | Dancing Robot indoor based on laser radar pedestrian tracting method | |
| CN109284786B (en) | SAR image terrain classification method for generating countermeasure network based on distribution and structure matching | |
| CN105137418B (en) | Multiple target tracking and data interconnection method based on complete adjacent fuzzy clustering | |
| CN104851097B (en) | The multichannel SAR GMTI methods aided in based on target shape and shade | |
| CN107255803A (en) | A kind of ship track logic start algorithm could | |
| CN105510896B (en) | A kind of weighted nearest neighbor numeric field data correlating method of centralization multi-radar data processing | |
| CN103268496B (en) | SAR image target recognition method | |
| CN108765470A (en) | One kind being directed to the improved KCF track algorithms of target occlusion | |
| CN106408591A (en) | Anti-blocking target tracking method | |
| CN110443138B (en) | Joint SVM and CNN multi-target classification method based on vehicle-mounted millimeter wave radar | |
| CN106842165A (en) | One kind is based on different distance angular resolution radar centralization asynchronous fusion method | |
| CN107301657B (en) | A kind of video target tracking method considering target movable information | |
| CN106304331A (en) | A kind of WiFi fingerprint indoor orientation method | |
| CN106443618B (en) | A kind of high resolution radar Data Association based on standard deviation ellipse parameter auxiliary | |
| CN112966571B (en) | Standing long jump flight height measurement method based on machine vision | |
| CN105120479A (en) | Signal strength difference correction method of Wi-Fi signals between terminals | |
| US20210223351A1 (en) | Detection Method and Detection Apparatus | |
| CN105223561A (en) | Based on the radar terrain object Discr. method for designing of space distribution | |
| CN105894022B (en) | A kind of adaptive layered association multi-object tracking method | |
| CN106291498A (en) | A kind of detecting and tracking combined optimization method based on particle filter | |
| CN111505649B (en) | Towed passive array sonar low signal-to-noise ratio ship moving target detection method | |
| CN110244294A (en) | A kind of correlating method of the metric data of multisensor | |
| CN106022250A (en) | Embryo splitting detection method based on cell movement information and gray property |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |