CN109239702B - Airport low-altitude flying bird number statistical method based on target state set - Google Patents
Airport low-altitude flying bird number statistical method based on target state set Download PDFInfo
- Publication number
- CN109239702B CN109239702B CN201811209544.3A CN201811209544A CN109239702B CN 109239702 B CN109239702 B CN 109239702B CN 201811209544 A CN201811209544 A CN 201811209544A CN 109239702 B CN109239702 B CN 109239702B
- Authority
- CN
- China
- Prior art keywords
- track
- bird
- target
- moment
- targets
- 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
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
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention provides a method for counting the number of low-altitude flying birds in an airport based on a target state set, which comprises the following steps: in each filtering period, performing interframe merging according to a plurality of previous continuous period bird detection radar echo measurement data to obtain a bird target candidate set, and removing existing targets in the bird target candidate set by using historical track information to obtain a new target candidate set of the filtering period; filtering the measurement data at the current moment according to the newly-generated target candidate set and the existing track at the previous moment to obtain a current target state set; and performing track connection and extinction judgment on the target state set, and outputting the number of the flying bird targets in the current monitoring range. The method effectively utilizes the historical track information, is suitable for airport low-altitude bird detection scenes with the rapidly changing number of the flying bird targets, and improves the estimation accuracy of the number of the flying bird targets.
Description
Technical Field
The invention relates to the technical field of low-altitude airspace security monitoring, in particular to radar image processing and target tracking, and specifically relates to an airport low-altitude flying bird number statistical method based on a target state set.
Background
The bird detection radar is widely applied to monitoring the condition of low-altitude birds in airports, can automatically run all day long without being influenced by factors such as weather and the like, and is an effective means for reducing bird strikes.
Due to the characteristics of strong flying randomness, weak radar echo and the like of low-altitude flying birds in an airport, stable flight tracks are difficult to form, meanwhile, the flying bird targets change rapidly in number, clutter interference is added, measurement of some targets may be missed at certain observation moments, some false alarm targets may be mixed, the number of the targets is difficult to accurately estimate by a traditional target tracking algorithm, and the tracking performance is reduced.
Disclosure of Invention
In order to overcome the problems, the invention provides a method for counting the number of low-altitude flying birds in an airport based on a target state set.
A method for counting the number of low-altitude flying birds in an airport based on a target state set comprises the following steps: step 1, in each filtering period, performing interframe merging according to a plurality of previous continuous period bird detection radar echo measurement data to obtain a bird target candidate set, and removing existing targets in the bird target candidate set by using historical track information to obtain a new target candidate set of the filtering period; step 2, filtering the measurement data at the current moment according to the newly-generated target candidate set and the track set at the previous moment to obtain a current target state set; step 3, combining the track set, the track state information matrix and the current target state set at the previous moment to obtain the track set and the track state information matrix at the current moment so as to realize track connection; and 4, judging the disappeared targets according to the track set and the track state information matrix at the current moment, and removing the targets to obtain the number of the bird targets in the current period.
Further, in step 1, the measurement data is in the form of two-dimensional coordinates, and the plurality of previous continuous periods are two previous connected periods.
Further, in step 1, the inter-frame merging specifically includes: the first inter-frame distance of any two measurements during two consecutive filtering cycles is calculated:
calculating a second inter-frame distance of any three measurements during three consecutive filtering cycles:
the following inter-frame merging conditions are satisfied
d1,ji∈[vmin,vmax]&&d1,pj∈[vmin,vmax]&&d2∈[0,amax]
Point pair (r)* (k,i),r* (k-1,j),r* (k-2,p)) Then, the new target is considered as a candidate new target, and r is recorded* (k,i)Calculating a candidate set of bird targets for the birth position of the candidate new target when measuring sets at the time of inputting k, k-1 and k-2, wherein T is a radar scanning period and r isk,iDenotes the i-th measured value, i 1,2, …, Q, generated at time k, which shares the Q-measured value, rk-1,jRepresents the j-th measured value generated at the time k-1, j being 1,2, …, N, the time being N measured values in total,rk-2,pdenotes the p-th measured value, p 1,2, …, L, occurring at the time k-2, which is shared by all L measured values, vmax、vmin、amaxRepresenting maximum, minimum flying speed and maximum acceleration of the bird, respectively.
Further, in step 2, a GM _ PHD filter is used for filtering.
Further, in step 3, the set of trajectories existing at time k-1 is denoted as Ek-1={ek-1,s}, s=1,2,…,n,ek-1,sRepresenting the s existing track value generated at the k-1 moment; the state information matrix corresponding to the track set is denoted as S1×nAnd the target state set obtained at the moment k is Xk={xk,i},i=1,2,…,m, xk,iRepresenting the ith effective target at the moment k, and establishing an initial mean vector set U ═ mu i1,2, n), wherein μi=ek-1,iFor each xk,jCalculate it and all mean vectors μiEuclidean distance of (d), mu, the smallest distanceiCorresponding trajectory ek-1,iIs xk,jIf a track e belongs tok-1,iIf there is no measurement value currently, a pre-estimation value is given according to the data of the first two frames.
Further, in step 4, the trace set E at time kk={ek,iH, where e is 1,2k,iThe value of the ith track at k time is shown, and a track state matrix S at k time is input1×tIn which S isiThe ith track e is recordedk,iThe current state, traverse the state matrix, if SiIf not less than 3, the ith track is proved to have disappeared from EkDeletion in ek,iFrom S1×tDeletion of Si。
The invention has the advantages that:
(1) the method for counting the number of the flying birds in the low altitude of the airport based on the target state set is suitable for the airport low altitude bird detection scene with the rapidly changing number of the flying birds targets, and the accuracy of estimating the number of the flying birds targets is improved;
(2) the method for counting the number of the low-altitude flying birds in the airport based on the target state set can effectively utilize historical track information and is not influenced by missed detection and false alarms.
Drawings
FIG. 1 is a flow chart of a statistical method of airport low-altitude bird counts for a target state set of the present invention;
FIG. 2 is a schematic diagram of a current frame filtered measurement position and target tracking according to an embodiment of the present invention;
FIG. 3 is a schematic diagram illustrating a measurement position and a target tracking condition after filtering a current frame according to an embodiment of the present invention;
fig. 4 is a diagram illustrating a target number estimation result after filtering a current frame according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples.
The invention provides a method for counting the number of low-altitude flying birds in an airport based on a target state set, and a flow chart of the method is shown in figure 1. Firstly, in each filtering period, carrying out frame-to-frame combination according to a plurality of previous continuous period bird detection radar echo measurement data to obtain a bird target candidate set, and removing existing targets in the bird target candidate set by using historical track information to obtain a new target candidate set of the filtering period; then, filtering the measurement data at the current moment according to the newly-generated target candidate set and the existing track at the previous moment to obtain a current target state set; and finally, performing track connection and casualty judgment on the target state set output by the filter, and outputting the target number of the flying birds in the current monitoring range.
The concrete description is as follows:
in order to detect a new bird target at time k, it is assumed that the measurement sets at times k, k-1 and k-2 are Mk={rk,i}、Mk-1={rk-1,j}、Mk-2={rk-2,pWherein r isk,i(i 1,2, …, Q) represents the i-th measured value generated at time k, which is the same as the Q-measured value, and r is the same as the Q-measured valuek-1,j(j-1, 2, …, N) represents the j-th measured value generated at the time k-1, which is shared by N measured values, rk-2,p(p-1, 2, …, L) represents the p-th measured value generated at the time k-2, which is the total timeThere were L measurements. In a two-dimensional scenario, the measurement value r ═ x, y represents the two-dimensional coordinates of a certain position.
Define any two measurement values the first inter-frame distance as:
the second inter-frame distance defining any three consecutive metrology values is:
wherein T is the radar scan period, r1、r2、r3Is any three-point measurement value. The physical meaning of the first inter-frame distance is the speed of assuming that the target makes uniform linear motion between two continuous measurement values, and the physical meaning of the second inter-frame distance is the acceleration of assuming that the target makes uniform acceleration linear motion between three continuous measurement values.
The first inter-frame distance of any two measurements during two consecutive filtering cycles is calculated:
a second inter-frame distance is calculated for any three measurements during three consecutive filtering cycles:
the following inter-frame merging conditions are satisfied
d1,ji∈[vmin,vmax]&&d1,pj∈[vmin,vmax]&&d2∈[0,amax]
Point pair (r)* (k,i),r* (k-1,j),r* (k-2,p)) Then it is considered as a candidate new target, and r is recorded* (k,i)Is the birth location of the candidate new-born target. Wherein v ismax、vmin、amaxThe maximum flying speed, the minimum flying speed and the maximum acceleration of the flying bird are respectively represented, and are defined as a combined threshold value of the first frame distance and the second frame distance, and can be obtained through statistics.
According to the method, the measurement set M of k, k-1 and k-2 time is inputk、Mk-1、Mk-2Calculating to obtain a candidate set of flying bird targets
The critical distance between any two measurement values is defined as:
where T is the radar scan period and r, e are two different measurement values.
Assume that the k-1 time history track set is Ek-1={ek-1,sIn which ek-1,s(s-1, 2, …, n) represents the s-th existing trace value generated at time k-1.
the following conditions of picking and merging are met:
point r of* (k,i)Forming a new target candidate set, and recording Bk as { r ═ r** (k,i)X, X indicates that X new targets are detected at time k, where X is equal to 1,2
Step 2, establishing a target state set:
the measurement set at the time k comprises a large number of clutters, the measurement set is filtered through the existing GM _ PHD filter, effective target measurement is extracted from the clutters, and a target state set is established. But the filter requires the birth information of the newborn target as a priori and is generally specified by human beings. However, in a low-altitude bird detection scene near an airport, the birth information of the target of the new-born bird cannot be acquired manually, and can be acquired by the algorithm in the step 1.
The newly generated target candidate set B detected in the step 1 is collectedkAnd the already existing trajectory set E at the time k-1k-1As initial information of the filter, and inputting the measurement M of k timekPerforming filtering to obtain a set of filtering results X at time kk={xk,i},i=1,2,…,m,xk,iIndicating the valid target position at time k.
Step 3, connecting tracks;
since the target state set established in step 2 is a disordered state set, a stable bird trajectory cannot be directly obtained, and thus trajectory linkage is performed.
Inputting the existing track set E at the moment of k-1k-1And a set of filtering results X of the k-time filterkAnd a state information matrix S of the k-1 time trace1×n(initialized to 0 at time k-0, followed by iterative update maintenance at each time) the following procedure is performed:
selection of Ek-1All values in (d) as initial mean vector U ═ μi}(i=1,2,...,n)//μi=ek-1,i
The updated track set E at the moment k can be obtainedk={e k,i1,2, t), wherein ek,i=CiAnd a trajectory state matrix S1×t。
And 4, casualty judgment:
the dead target can not be measured any more, although some tracked targets may have missed detection in a certain detection, because the track set and the track state matrix record the continuous missed detection times of a certain target, statistically, if the continuous missed detection times of a certain target are more than 3, the target can be considered to be dead.
Inputting a k time trajectory set Ek={ek,iH (i 1, 2.., t), a trajectory state matrix S at time k is input1×tThe following algorithm is executed:
the current track set E can be obtainedk={e k,i1, 2.. times.m) (m ≦ t). The extinction track can be detected in 3 periods by detecting the track state matrix. If the false alarm exists, the measurement of the false alarm cannot be continuously updated, and the false alarm can be quickly detected by the extinction detection algorithm, so that the aim of reducing the false alarm is fulfilled. The target number at the time k is finally obtained as m.
The method of the present invention is applied to illustrate how to count the number of low-altitude birds in an airport with reference to fig. 2 to 4.
it is assumed that the bird target information in the detection range is shown in the following table:
|
1 | 2 | 3 | 4 |
Birth site X | -67.1709 | 795.0736 | -674.6922 | -620.0959 |
Birth position Y | -7.4683 | 446.4412 | -470.3579 | -839.2658 |
Time of |
1 | 1 | 38 | 38 |
Moment of death | 37 | 60 | 80 | 80 |
The object numbers correspond to the object numbers in fig. 2. The current scan period k is 40. It is known thatThe measurement numbers k 38, k 39, and k 40 are respectively 38, 39, and 40, and the filtering result set of k 39 obtained from the previous scanning cycle is { (-410.9456, -600.5789), (832.1801,647.2697) }, corresponding to the last two points of the track No. 1 and 2 in fig. 2, two targets are found, the set records the positions of the targets, and the corresponding track state matrix S is [2,0 ═ 2]. Selecting vmin=0,vmax=50,amax20. The new target candidate set at the time k-40 can be obtained by the method in step 1 as follows: { (-670.1934, -489.4508), (-611.3860, -813.9683) }, which corresponds to just two points A, B in fig. 2, it can be seen that targets # 3 and # 4 were successfully detected at time 40.
Step 2, establishing a target state set:
the newly created target candidate information detected in step 1 and the existing trajectory information at time 39 are used as initial information of the filter, and the measurement at time 40 is input to perform filtering. The set of filtering results at time k-40 is obtained as: { (827.0614,653.6661), (-670.1934, -489.4508), (-611.3860, -813.9683) }, which corresponds exactly to point C, A, B in fig. 3.
Step 3, track connection:
when the 39-time filter result set { (-410.9456, -600.5789), (832.1801,647.2697) }, the corresponding trajectory state matrix S { (2, 0], and the 40-time filter result set { (827.0614,653.6661), (-670.1934, -489.4508), (-611.3860, -813.9683) }, are input into the trajectory connection algorithm described above, the 40-time filter result { (-415.5473, -608.1287), (827.0614,653.6661), (-670.1934, -489.4508), (-611.3860, -813.9683) }, which exactly corresponds to point D, C, A, B in fig. 3, and the corresponding trajectory state matrix S [3,0,0,0] can be obtained.
And 4, casualty judgment:
after the above extinction judgment algorithm is executed, it can be known that the target No. 1 has been extinguished, and the filter trajectory set at 40 moments is obtained as { (827.0614,653.6661), (-670.1934, -489.4508), (-611.3860, -813.9683) }, and the corresponding trajectory state matrix is S ═ 0,0, 0. The target number at time 40 is determined to be 3 as indicated by point E in fig. 4.
Because the bird flight track is regular, such as a straight line or an arc line, the historical track information can be used for effectively eliminating false alarms and predicting missed targets, and the accuracy of quantity estimation is improved.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (5)
1. A method for counting the number of low-altitude flying birds in an airport based on a target state set comprises the following steps:
step 1, in each filtering period, performing frame-to-frame merging according to a plurality of previous continuous period bird detection radar echo measurement data to obtain a bird target candidate set, and removing existing targets in the bird target candidate set by using historical track information to obtain a new target candidate set of the filtering period;
step 2, filtering the measurement data at the current moment according to the new target candidate set and the track set at the previous moment to obtain a current target state set;
step 3, combining the track set, the track state information matrix and the current target state set at the previous moment to obtain the track set and the track state information matrix at the current moment so as to realize track connection;
step 4, judging the disappeared targets according to the track set and the track state information matrix at the current moment, removing the targets to obtain the number of the flying bird targets in the current period,
in step 1, the inter-frame merging specifically includes calculating a first inter-frame distance of any two measurement values during two consecutive filtering cycles:
a second inter-frame distance is calculated for any three measurements during three consecutive filtering cycles:
the following inter-frame merging conditions are satisfied
d1,ji∈[vmin,vmax]&&d1,pj∈[vmin,vmax]&&d2∈[0,amax]Point pairThen the new target is considered as a candidate new target and is recordedInputting measurement sets at k, k-1 and k-2 moments for the birth position of the candidate new target, and calculating to obtain a bird target candidate set, wherein T is a radar scanning period, r is(k,i)Denotes the i-th measured value, i 1,2, …, Q, generated at time k, which shares the Q-measured value, r(k-1,j)Represents the j-th measured value generated at the time k-1, j being 1,2, …, N, and N being the total measured value of N, r(k-2,p)Denotes the p-th measured value, p 1,2, …, L, occurring at the time k-2, which shares the L measured value, vmax、vmin、amaxRepresenting maximum, minimum flying speed and maximum acceleration of the bird, respectively.
2. The method according to claim 1, wherein in step 1, the measured data is in the form of two-dimensional coordinates, and the plurality of previous consecutive periods are two previous consecutive periods.
3. The method according to claim 1, wherein in step 2, a GM-PHD filter is used for filtering.
4. The method according to claim 1, wherein in step 3, the existing trajectory set at the time k-1 is denoted as Ek-1={ek-1,s},s=1,2,...,n,ek-1,sRepresenting the s existing track value generated at the k-1 moment; the state information matrix corresponding to the track set is denoted as Sl×nAnd the target state set obtained at the moment k is Xk={xk,i},i=1,2,...,m,xk,iRepresenting the ith effective target at the moment k, and establishing an initial mean vector set U ═ mui1,2, n, where μi=ek-1,iFor each xk,jCalculate it and all mean vectors μiOf the Euclidean distance, mu, of the smallest distanceiCorresponding trajectory ek-1,iIs xk,jIf a track e belongs tok-1,iIf there is no measurement value currently, a pre-estimation value is given according to the data of the first two frames.
5. The method for counting the number of low-altitude flying birds in airport according to claim 1, wherein in step 4, the trace set E at time k isk={ek,i1,2, t, wherein ek,iThe value of the ith track at the k moment is shown, and a track state matrix S at the k moment is inputl×tIn which S isiThe ith track e is recordedk,iThe current state, traverse the state matrix, if SiIf not less than 3, the ith track is proved to have disappeared from EkDelete in ek,iFrom Sl×tDeletion of Si。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811209544.3A CN109239702B (en) | 2018-10-17 | 2018-10-17 | Airport low-altitude flying bird number statistical method based on target state set |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811209544.3A CN109239702B (en) | 2018-10-17 | 2018-10-17 | Airport low-altitude flying bird number statistical method based on target state set |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109239702A CN109239702A (en) | 2019-01-18 |
CN109239702B true CN109239702B (en) | 2022-07-19 |
Family
ID=65053449
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811209544.3A Active CN109239702B (en) | 2018-10-17 | 2018-10-17 | Airport low-altitude flying bird number statistical method based on target state set |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109239702B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110031816B (en) * | 2019-03-22 | 2021-04-27 | 中国民航科学技术研究院 | Airport flight area non-cooperative target classification and identification method based on bird detection radar |
CN110188832B (en) * | 2019-05-08 | 2020-12-08 | 中国民航科学技术研究院 | Airport bird-remaining activity rhythm statistical analysis method based on bird-detecting radar |
CN111812635B (en) * | 2020-06-10 | 2023-08-01 | 广东电网有限责任公司 | Bird flight trajectory prediction method based on anti-shake filtering method weighting |
CN115880646A (en) * | 2023-02-20 | 2023-03-31 | 中国民航大学 | Method for identifying in-out-of-position state of airplane |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101697007A (en) * | 2008-11-28 | 2010-04-21 | 北京航空航天大学 | Radar image-based flyer target identifying and tracking method |
US8427359B1 (en) * | 2011-01-06 | 2013-04-23 | Sandia Corporation | Tracking moving radar targets with parallel, velocity-tuned filters |
JP2014055883A (en) * | 2012-09-13 | 2014-03-27 | Furuno Electric Co Ltd | Radar system, detection method and detection program |
DK201300589A1 (en) * | 2013-10-15 | 2015-04-27 | Robin Radar Facilities Bv | Dynamic alarm zones for bird detection systems |
CN105548970A (en) * | 2015-12-11 | 2016-05-04 | 无锡市雷华科技有限公司 | Flying bird detection radar processor |
JP2016090409A (en) * | 2014-11-06 | 2016-05-23 | いであ株式会社 | Bird observation method |
CN106546975A (en) * | 2016-10-14 | 2017-03-29 | 中国民航科学技术研究院 | A kind of small-sized unmanned plane based on radar data and flying bird classifying identification method |
CN107292265A (en) * | 2017-06-20 | 2017-10-24 | 中国电子科技集团公司第二十八研究所 | A kind of target trajectory rapid extracting method based on motor-driven detection |
-
2018
- 2018-10-17 CN CN201811209544.3A patent/CN109239702B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101697007A (en) * | 2008-11-28 | 2010-04-21 | 北京航空航天大学 | Radar image-based flyer target identifying and tracking method |
US8427359B1 (en) * | 2011-01-06 | 2013-04-23 | Sandia Corporation | Tracking moving radar targets with parallel, velocity-tuned filters |
JP2014055883A (en) * | 2012-09-13 | 2014-03-27 | Furuno Electric Co Ltd | Radar system, detection method and detection program |
DK201300589A1 (en) * | 2013-10-15 | 2015-04-27 | Robin Radar Facilities Bv | Dynamic alarm zones for bird detection systems |
JP2016090409A (en) * | 2014-11-06 | 2016-05-23 | いであ株式会社 | Bird observation method |
CN105548970A (en) * | 2015-12-11 | 2016-05-04 | 无锡市雷华科技有限公司 | Flying bird detection radar processor |
CN106546975A (en) * | 2016-10-14 | 2017-03-29 | 中国民航科学技术研究院 | A kind of small-sized unmanned plane based on radar data and flying bird classifying identification method |
CN107292265A (en) * | 2017-06-20 | 2017-10-24 | 中国电子科技集团公司第二十八研究所 | A kind of target trajectory rapid extracting method based on motor-driven detection |
Non-Patent Citations (4)
Title |
---|
"Classification of radar targets using synthetic neural networks";Jouny, I 等;《IEEE Transactions on Aerospace and Electronic Systems》;19931230;第29卷(第2期);正文第336-344页 * |
"Multiple hypothesis tracking for multiple target tracking";S.S. Blackman;《Aerospace and Electronic Systems Magazine》;20041230;第19卷(第1期);正文第5-18页 * |
基于鸟类目标散射特性分析的雷达探鸟实验;陈唯实等;《航空学报》;20090725(第07期);正文第154-160页 * |
雷达探鸟技术发展与应用综述;陈唯实等;《现代雷达》;20170215(第02期);正文第7-16页 * |
Also Published As
Publication number | Publication date |
---|---|
CN109239702A (en) | 2019-01-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109239702B (en) | Airport low-altitude flying bird number statistical method based on target state set | |
CN108986064B (en) | People flow statistical method, equipment and system | |
CN109522854B (en) | Pedestrian traffic statistical method based on deep learning and multi-target tracking | |
KR20200006167A (en) | Vessel automatic tracking method and system based on deep learning network and average movement | |
CN110542885B (en) | Millimeter wave radar target tracking method in complex traffic environment | |
WO2012009947A1 (en) | Device and method of signature-driven multi-target tracking | |
CN104574439A (en) | Kalman filtering and TLD (tracking-learning-detection) algorithm integrated target tracking method | |
CN105761276B (en) | Based on the iteration RANSAC GM-PHD multi-object tracking methods that adaptively newborn target strength is estimated | |
CN108537829B (en) | Monitoring video personnel state identification method | |
CN103413324A (en) | Automatic target tracking method for aerially photographed videos | |
CN107436434B (en) | Track starting method based on bidirectional Doppler estimation | |
CN113030973B (en) | Scene monitoring radar signal processing system and method | |
CN112396626A (en) | Sequence image population identification method for monitoring migratory bird migration | |
CN111259332B (en) | Fuzzy data association method and multi-target tracking method in clutter environment | |
CN116778290A (en) | Radar vision data association method based on deep learning algorithm | |
Yi et al. | Multi-Person tracking algorithm based on data association | |
CN111898746B (en) | Deep learning method for continuous relevance of broken flight path | |
Luber et al. | Better models for people tracking | |
CN113283330A (en) | Video SAR moving target detection method based on deep learning and multi-target tracking algorithm | |
CN110889347B (en) | Density traffic flow counting method and system based on space-time counting characteristics | |
CN106128105B (en) | A kind of traffic intersection pedestrian behavior monitoring system | |
CN116736290A (en) | Multi-target tracking method based on air biological motion situation feature assistance | |
Roopchand et al. | Bat detection and tracking toward batsman stroke recognition | |
CN116008936A (en) | Human body track tracking detection method based on millimeter wave radar | |
CN112711025A (en) | Underwater multi-station combined multi-target tracking method and system |
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 |