WO2022116375A1 - 高分辨传感器弱小多目标检测前跟踪方法 - Google Patents
高分辨传感器弱小多目标检测前跟踪方法 Download PDFInfo
- Publication number
- WO2022116375A1 WO2022116375A1 PCT/CN2021/072857 CN2021072857W WO2022116375A1 WO 2022116375 A1 WO2022116375 A1 WO 2022116375A1 CN 2021072857 W CN2021072857 W CN 2021072857W WO 2022116375 A1 WO2022116375 A1 WO 2022116375A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- track
- target
- tentative
- confirmed
- targets
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000001514 detection method Methods 0.000 claims abstract description 32
- 238000005259 measurement Methods 0.000 claims abstract description 20
- 230000004927 fusion Effects 0.000 claims abstract description 18
- 239000002245 particle Substances 0.000 claims abstract description 16
- 238000001914 filtration Methods 0.000 claims abstract description 7
- 238000009826 distribution Methods 0.000 claims description 5
- 238000012423 maintenance Methods 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 2
- 238000012790 confirmation Methods 0.000 claims description 2
- 238000007499 fusion processing Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000015556 catabolic process Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000009827 uniform distribution Methods 0.000 description 1
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
- G01S13/66—Radar-tracking systems; Analogous systems
Definitions
- the invention relates to a target detection and tracking method, in particular to a weak and small multi-target detection and tracking method, which belongs to the field of information fusion and is suitable for the detection and tracking of multiple weak and small targets by high-resolution sensors.
- High-resolution sensors can provide more target structure information, and have an extremely broad application background in target detection, tracking and recognition, and are one of the important development directions of modern sensor technology.
- the target echo usually occupies multiple resolution units.
- the target signal-to-noise ratio is high, the target can be achieved by means of target recognition and building a suitable target expansion model.
- the number of corrections, and the estimated target scattering center has improved the target tracking accuracy; however, when the target signal-to-noise ratio is low, the target echo is usually submerged in the noise, and the target measurement data is sometimes absent.
- the existing method It is difficult to detect it effectively, let alone identify and model the target. Therefore, how to realize the effective detection and tracking of multiple weak and small targets with high-resolution sensors is a difficult problem at present.
- TBD Tracking before detection
- the weak and small target detection and tracking method based on TBD technology has two obvious defects: (1) This method treats the target as a "point target", which will cause a target expanded in multiple resolution units to be detected as multiple targets, and then As a result, the estimation of the number of targets is obviously excessive, and the algorithm complexity and calculation increase significantly; (2) When the target is missed, the particle swarm representing the target in the filter degrades rapidly, resulting in the easy loss of the target at subsequent times.
- the purpose of the present invention is to propose a high-resolution sensor pre-detection tracking method for weak and small targets, so as to solve the problem that the target number estimation is obviously excessive and the algorithm complexity caused by the weak and small target detection and tracking method based on TBD technology treating the target as a "point target".
- the particle swarm representing the target in the filter is rapidly degraded, and the target is easily lost at subsequent moments.
- a main filter is constructed, and the initialization of the filter is performed.
- a missed detection target set is constructed and initialized as an empty set.
- Step 2 Sensor measurement acquisition
- Step 3 Generation of Missing Target Aided Search Particle Sets
- the missed target set is an empty set, go to step 4 directly; otherwise, first perform a one-step prediction on any target in the missed target set to obtain the predicted value of the target state at the current moment, and then use the predicted value as the mean value and the target measurement
- the error covariance is the variance
- a multi-dimensional normal distribution is constructed, and the auxiliary search particle set used to assist the search for the missed target is sampled from the normal distribution, and finally the auxiliary search particle set of all the missed targets is added to the main filter, complete The main filter is corrected and the set of missed targets is reset to the empty set.
- Step 6 Target Track Management
- Step 7 Go to the next moment and repeat steps 2 to 7 until the sensor is turned off.
- step 5 is specifically:
- the ring gate only falls within a certain track head, the state estimate is considered to be derived from the track head, and it is classified as the "homologous" target of the track head;
- the movement direction of the target is calculated according to the target velocity information contained in the state estimation, and it is divided into the confirmation track with the smallest angle between the target heading and its movement direction, "Homologous" targets of tentative, probable, or track headers;
- step 6 is specifically:
- the high-resolution sensor pre-detection tracking method for weak and small multi-targets proposed by the present invention fuses the estimated values from the same target, which solves the obvious false estimation of the number of targets caused by the background technology treating the target as a "point target". Many problems such as algorithm complexity and calculation increase significantly.
- the present invention can effectively prevent the problem of rapid degradation of particle swarms representing missed detections in the main filter by constructing an auxiliary search mechanism for missed targets in the main filter, thereby solving the problem that the target is easily lost in subsequent moments when the target is missed. The problem.
- FIG. 1 is a schematic diagram of the overall flow of the tracking method before detection of weak and small multi-targets by a high-resolution sensor of the present invention.
- FIG. 2 is a schematic diagram of the “same source” target division of the high-resolution sensor tracking method before weak and small multi-target detection of the present invention.
- FIG. 3 is a detection and tracking result diagram of the high-resolution sensor tracking method before weak and small multi-target detection of the present invention.
- FIG. 4 is a schematic diagram of track maintenance and upgrade of the high-resolution sensor tracking method before weak and small multi-target detection of the present invention.
- FIG. 5 is a schematic diagram of track degradation and cancellation of the tracking method before detection of weak and small multi-targets by high-resolution sensors of the present invention.
- the total simulation time is 80s, and the PHD filter is implemented by particle filtering.
- the measurement noise follows a Gaussian distribution with zero mean, that is, (x k , y k )
- the influence of the target with the intensity I k at the position of (x k , y k ) on the resolution unit (i, j) can be approximated by the divergence function as
- ⁇ is a known parameter, representing the number of blurred spots in the sensor, and the measurement at time k can be expressed as
- the target can appear and disappear randomly in the monitoring area, and its appearance position conforms to a uniform distribution in the monitoring area.
- w zero mean Gaussian white noise, and its covariance matrix is
- Step 1 Carry out system initialization according to the method described in Step 1 of the Summary of the Invention
- Step 2 Perform sensor measurement and acquisition according to the method described in Step 2 of the Summary of the Invention
- Step 3 According to the method described in Step 3 of the section of the content of the invention, the auxiliary search particle set of the missed target is generated;
- Step 4 Perform measurement coarse filtering according to the method described in Step 4 of the Summary of the Invention
- Step 5 According to the method described in Step 5 of the Summary of the Invention, the division and fusion of "homologous" targets are carried out;
- Step 6 Carry out target track management according to the method described in Step 6 of the Summary of the Invention
- Step 7 According to the method described in Step 7 of the Summary of the Invention, go to the next moment, and repeat Steps 2 to 7 until the sensor is turned off.
- the high-resolution sensor tracking method before weak and small multi-target detection proposed by the present invention effectively divides and fuses "same source” targets (see step 5), and fuses the estimated values from the same target ( See accompanying drawing 2), has realized the more accurate estimation (see accompanying drawing 3) to the target quantity and the target state at each moment, therefore solves the background technology to treat the target as "point target” to cause the target number estimate to be obviously too many
- the auxiliary search mechanism for missed targets in the main filter see steps 3 and 6 and accompanying drawings 4 and 5
- it can effectively prevent the filter
- the particle swarm representing the target in the target degenerates rapidly, so as to solve the problem that the target is easily lost in the subsequent moments caused by the missed detection of the target.
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)
Abstract
一种高分辨传感器弱小多目标检测前跟踪方法,该方法包括:步骤1:***初始化;步骤2:传感器量测获取;步骤3:漏检目标辅助搜索粒子集生成;步骤4:量测粗滤波;步骤5:"同源"目标划分与融合;步骤6:目标航迹管理;步骤7:转下一时刻,重复步骤2~步骤7直至传感器关机。该方法通过对"同源"目标进行有效划分与融合,将源于同一目标的估计值进行融合处理,并通过在主滤波器中构建漏检目标辅助搜索机制,克服了基于TBD的弱小目标检测跟踪方法的局限性,具有较强的工程应用价值和推广前景。
Description
本发明涉及一种目标检测跟踪方法,特别是涉及一种弱小多目标的检测跟踪方法,属于信息融合领域,适应于高分辨传感器对多个弱小目标的检测与跟踪。
高分辨传感器可提供更多的目标结构信息,在目标检测、跟踪与识别等方面具有极其广阔的应用背景,是现代传感器技术的重要发展方向之一。但是,由于该类传感器分辨单元的尺寸远小于目标尺寸,目标回波通常会占据多个分辨单元,当目标信噪比较高时,可通过目标识别并构建合适的目标扩展模型等手段实现目标个数的修正,并估计出目标散射中心已提高目标跟踪精度;然而,当目标信噪比较低时,目标回波通常被淹没在噪声里,目标量测数据时有时无,现有的方法很难对其进行有效检测,更无法对目标进行识别和建模。因此,如何实现高分辨传感器对多个弱小目标的有效检测和跟踪是当前的难点问题。
检测前跟踪(TBD)技术是一种目前最常用且行之有效的弱小的检测和跟踪方法,该方法以时间为代价,通过长时间的积累提高目标的信噪比,进而实现对弱小目标的有效检测和跟踪。现有的基于TBD技术的弱小目标检测跟踪方法主要通过以下步骤实现:
1)存储多帧未经门限处理或者较低门限处理的量测数据;
2)按照假定的目标运动轨迹对量测数据进行有效积累;
3)宣告目标存在的同时提取出目标的轨迹。
基于TBD技术的弱小目标检测跟踪方法具有两个明显的缺陷:(1)该方法将目标当成“点目标”来处理,会导致扩展在多个分辨单元的一个目标被检测成多个目标,进而导致目标个数估计明显虚多、算法复杂度和计算显著增大;(2)目标出现漏检时,滤波器中代表该目标的粒子群迅速退化,造成后续时刻目标容易的丢失。
发明内容
本发明的目的是提出一种高分辨传感器弱小多目标检测前跟踪方法,解决基于TBD技术的弱小目标检测跟踪方法将目标当成“点目标”来造成的目标个数估计明显虚多、算法复杂度和计算显著增大,以及目标出现漏检时,滤波器中代表该目标的粒子群迅速退化造成的后续时刻目标容易丢失等问题。
本发明提出的高分辨传感器弱小多目标检测前跟踪方法的技术方案包括以下步骤:
步骤1:***初始化
基于概率假设密度滤波构建一个主滤波器,并进行滤波器的初始化处理,另外构建一个 漏检目标集并初始化为空集。
步骤2:传感器量测获取
获取高分辨传感器当前时刻的非门限处理的弱小多目标量测数据,并进行A/D转换,送数据处理机;
步骤3:漏检目标辅助搜索粒子集生成
若漏检目标集为空集,直接转步骤4;否则,首先对漏检目标集中的任意目标进行一步预测,得到当前时刻该目标状态的预测值,然后以该预测值为均值、目标量测误差协为方差,构造多维正态分布,并从该正态分布采样用于辅助搜索该漏检目标的辅助搜索粒子集,最后将所有漏检目标的辅助搜索粒子集添加至主滤波器,完成主滤波器修正,并将漏检目标集重置为空集。
步骤4:量测粗滤波
利用修正后的主滤波器对量测数据进行滤波,得到粗略的目标个数及状态估计,其中状态估计包含了对目标的位置和速度估计;
步骤5:“同源”目标划分与融合
综合利用各目标的航迹对粗略的目标状态估计进行“同源”目标的划分与融合,修正目标个数和目标状态估计;
步骤6:目标航迹管理
对确认航迹、暂定航迹、可能航迹及航迹头进行航迹维持、升级、降级及撤销操作,并输出当前时刻的各目标航迹信息;
步骤7:转下一时刻,重复步骤2~步骤7直至传感器关机。
具体的,所述步骤5具体为:
1)对前一时刻已存在的各确定航迹、暂定航迹及可能航迹进行一步预测,得到各航迹的预测值,并依据预测协方差建立各自的椭圆波门,对前一时刻已存在的航迹头,根据目标最小和最大可能速度建立环形波门;
2)对任意粗略的目标状态估计,若该状态估计
(1)未落入任何椭圆波门或环形波门,将该状态估计定义为航迹头;
(2)仅落入某一确认航迹的椭圆波门,则认为该状态估计是源于该确认航迹,并将其划分为该确认航迹的“同源”目标;
(3)仅落入某一暂定航迹的椭圆波门,则认为该状态估计是源于该暂定航迹,并将其划分为该暂定航迹的“同源”目标;
(4)仅落入某一可能航迹的椭圆波门,则认为该状态估计是源于该可能航迹,并将其划分为该可能航迹的“同源”目标;
(5)仅落入某一航迹头的环形波门,则认为该状态估计是源于该航迹头,并将其划分为该航迹头的“同源”目标;
(6)同时落入多个椭圆波门或环形波门,则根据该状态估计包含的目标速度信息计算目标的运动方向,并将其划分为目标航向与其运动方向夹角最小的确认航迹、暂定航迹、可能航迹或者航迹头的“同源”目标;
3)利用概率数据互联的方法对各确认航迹、暂定航迹、可能航迹和航迹头的“同源”目标进行加权融合,得到当前时刻各自的融合目标状态估计。
具体的,所述步骤6具体为:
1)航迹维持与升级
(1)对任意确认航迹,以其融合目标状态估计更新并输出该确认航迹,实现航迹的维持;
(2)对任意可能航迹,以其融合目标状态估计更新该可能航迹,然后将该可能航迹升级为确认航迹并输出;
(3)对任意暂定航迹,以其融合目标状态估计更新该暂定航迹,然后将该暂定航迹升级为确认航迹并输出;
(4)对任意航迹头,将其融合目标状态估计与之关联,形成可能航迹;
2)航迹降级与撤销
(1)对当前时刻未获得任何“同源”目标的任意确认航迹,以该确认航迹的预测值更新该确认航迹,并将该确认航迹降级为暂定航迹,将该目标添加至漏检目标集;
(2)对当前时刻未获得任何“同源”目标的任意暂定航迹予以撤销;
(3)对当前时刻未获得任何“同源”目标的任意可能航迹,以其预测值更新对该可能航迹进行更新,并将可能航迹降级为暂定航迹;
(4)对当前时刻未获得任何“同源”目标的任意航迹头予以撤销。
和背景技术相比,本发明的有益效果说明:
1)本发明提出的高分辨传感器弱小多目标检测前跟踪方法,将源于同一目标的估计值进行融合处理,解决了背景技术将目标当成“点目标”来处理造成的目标个数估计明显虚多、算法复杂度和计算显著增大等问题。
2)本发明通过在主滤波器中构建漏检目标辅助搜索机制,可有效防止了主滤波器中代表漏检的粒子群迅速退化的问题,从而解决了目标出现漏检时后续时刻目标容易丢失的问题。
附图1是本发明的高分辨传感器弱小多目标检测前跟踪方法的整体流程示意图。
附图2是本发明的高分辨传感器弱小多目标检测前跟踪方法的“同源”目标划分示意图。
附图3是本发明的高分辨传感器弱小多目标检测前跟踪方法的检测跟踪结果图。
附图4是本发明的高分辨传感器弱小多目标检测前跟踪方法的航迹维持与升级示意图。
附图5是本发明的高分辨传感器弱小多目标检测前跟踪方法的航迹降级与撤销示意图。
不失一般性,假设一个二维的多目标跟踪场景,总仿真时间为80s,PHD滤波器通过粒子滤波实现。传感器位于坐标原点,采样周期T=1s,每时刻可提供的某个检测区域内的二维图像,每张图像包含m×n=20×20个分辨单元,每一个分辨单元的尺度设置为Δ
x=Δ
y=1,以
表示时刻k分辨单元(i,j)的观测数据
其中
为目标对分辨单元(i,j)强度的影响,
为分辨单元(i,j)的量测噪声,并假设分辨单元与分辨单元、前一时刻与后一时刻之间的量测噪声相互独立。为简化起见,假设量测噪声服从零均值的高斯分布,也就是
(x
k,y
k)位置强度为I
k的目标对分辨单元(i,j)的影响可以用散度函数近似表示为
∑为已知参数,表示传感器模糊斑点数量,k时刻的量测可以表述为
量测噪声方差设为σ=3.25,目标信号强度I=20,参数∑=0.7,根据信噪比公式
目标可在监测区域内随机的出现和消失,其出现的位置在监测区域内符合均匀分布,目标新生概率b
k=0.01,目标持续存在概率e
k=0.99,目标运动符合匀速直线运动模型
x
k+1=F(x
k)+w
k
其中
w为零均值高斯白噪声,其协方差矩阵为
其中
q
1和q
2分别表示目标运动状态和目标强度的过程噪声,设置为q
1=0.001,q
2=0.01。
滤波器参数设置:代表1个目标的粒子数L
0=3000,搜索新目标的粒子数J
k=4000,搜索1个漏检目标的粒子数S
k=1000,初步检测门限γ=2。
下面结合附图,对本发明的高分辨传感器弱小多目标检测前跟踪方法进行详细描述。
步骤1:按发明内容部分步骤1所述的方法进行***初始化;
步骤2:按发明内容部分步骤2所述的方法进行传感器量测获取;
步骤3:按发明内容部分步骤3所述的方法进行漏检目标辅助搜索粒子集生成;
步骤4:按发明内容部分步骤4所述的方法进行量测粗滤波;
步骤5:按发明内容部分步骤5所述的方法进行“同源”目标划分与融合;
步骤6:按发明内容部分步骤6所述的方法进行目标航迹管理;
步骤7:按发明内容部分步骤7所述的方法转下一时刻,重复步骤2~步骤7直至传感器关机。
实施例条件中,本发明提出的高分辨传感器弱小多目标检测前跟踪方法,通过对“同源”目标进行有效划分与融合(见步骤5),将源于同一目标的估计值进行融合处理(见附图2),实现了对目标数量和各时刻目标状态的较准确估计(见附图3),因此解决了背景技术将目标当成“点目标”来处理造成的目标个数估计明显虚多、算法复杂度和计算显著增大等问题;同时,通过在主滤波器中构建漏检目标辅助搜索机制(见步骤3和步骤6及附图4和附图5),可有效防止了滤波器中代表该目标的粒子群迅速退化,从而解决目标出现漏检造成的后续时刻目标容易丢失的问题。
Claims (3)
- 高分辨传感器弱小多目标检测前跟踪方法,其特征包括以下步骤:步骤1:***初始化基于概率假设密度滤波构建一个主滤波器,并进行滤波器的初始化处理,另外构建一个漏检目标集并初始化为空集。步骤2:传感器量测获取获取高分辨传感器当前时刻的非门限处理的弱小多目标量测数据,并进行A/D转换,送数据处理机;步骤3:漏检目标辅助搜索粒子集生成若漏检目标集为空集,直接转步骤4;否则,首先对漏检目标集中的任意目标进行一步预测,得到当前时刻该目标状态的预测值,然后以该预测值为均值、目标量测误差协为方差,构造多维正态分布,并从该正态分布采样用于辅助搜索该漏检目标的辅助搜索粒子集,最后将所有漏检目标的辅助搜索粒子集添加至主滤波器,完成主滤波器修正,并将漏检目标集重置为空集。步骤4:量测粗滤波利用修正后的主滤波器对传感器量测数据进行滤波,得到粗略的目标个数及状态估计,其中状态估计包含了对目标的位置和速度估计;步骤5:“同源”目标划分与融合综合利用各目标的航迹对粗略的目标状态估计进行“同源”目标的划分与融合,修正目标个数和目标状态估计;步骤6:目标航迹管理对确认航迹、暂定航迹、可能航迹及航迹头进行航迹维持、升级、降级及撤销操作,并输出当前时刻的各目标航迹信息;步骤7:转下一时刻,重复步骤2~步骤7直至传感器关机。
- 权利要求1所述的高分辨传感器弱小多目标检测前跟踪方法,其特征在于,所述步骤5具体为:1)对前一时刻已存在的各确定航迹、暂定航迹及可能航迹进行一步预测,得到各航迹的预测值,并依据预测协方差建立各自的椭圆波门,对前一时刻已存在的航迹头,根据目标最小和最大可能速度建立环形波门;2)对任意粗略的目标状态估计,若该状态估计(1)未落入任何椭圆波门或环形波门,将该状态估计定义为航迹头;(2)仅落入某一确认航迹的椭圆波门,则认为该状态估计是源于该确认航迹,并将其划分为该确认航迹的“同源”目标;(3)仅落入某一暂定航迹的椭圆波门,则认为该状态估计是源于该暂定航迹,并将其划分为该暂定航迹的“同源”目标;(4)仅落入某一可能航迹的椭圆波门,则认为该状态估计是源于该可能航迹,并将其划分为该可能航迹的“同源”目标;(5)仅落入某一航迹头的环形波门,则认为该状态估计是源于该航迹头,并将其划分为该航迹头的“同源”目标;(6)同时落入多个椭圆波门或环形波门,则根据该状态估计包含的目标速度信息计算目标的运动方向,并将其划分为目标航向与其运动方向夹角最小的确认航迹、暂定航迹、可能航迹或者航迹头的“同源”目标;3)利用概率数据互联的方法对各确认航迹、暂定航迹、可能航迹和航迹头的“同源”目标进行加权融合,得到当前时刻各自的融合目标状态估计。
- 利要求1所述的高分辨传感器弱小多目标检测前跟踪方法,其特征在于,所述步骤6具体为:1)航迹维持与升级(1)对任意确认航迹,以其融合目标状态估计更新并输出该确认航迹,实现航迹的维持;(2)对任意可能航迹,以其融合目标状态估计更新该可能航迹,然后将该可能航迹升级为确认航迹并输出;(3)对任意暂定航迹,以其融合目标状态估计更新该暂定航迹,然后将该暂定航迹升级为确认航迹并输出;(4)对任意航迹头,将其融合目标状态估计与之关联,形成可能航迹;2)航迹降级与撤销(1)对当前时刻未获得任何“同源”目标的任意确认航迹,以该确认航迹的预测值更新该确认航迹,并将该确认航迹降级为暂定航迹,将该目标添加至漏检目标集;(2)对当前时刻未获得任何“同源”目标的任意暂定航迹予以撤销;(3)对当前时刻未获得任何“同源”目标的任意可能航迹,以其预测值更新对该可能航迹进行更新,并将可能航迹降级为暂定航迹;(4)对当前时刻未获得任何“同源”目标的任意航迹头予以撤销。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011383508.6 | 2020-12-01 | ||
CN202011383508.6A CN112649802B (zh) | 2020-12-01 | 2020-12-01 | 高分辨传感器弱小多目标检测前跟踪方法 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022116375A1 true WO2022116375A1 (zh) | 2022-06-09 |
Family
ID=75351063
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2021/072857 WO2022116375A1 (zh) | 2020-12-01 | 2021-01-20 | 高分辨传感器弱小多目标检测前跟踪方法 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN112649802B (zh) |
WO (1) | WO2022116375A1 (zh) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115099343A (zh) * | 2022-06-24 | 2022-09-23 | 江南大学 | 一种有限视野下分布式标签多伯努利融合跟踪方法 |
CN115407299A (zh) * | 2022-09-15 | 2022-11-29 | 中国人民解放军国防科技大学 | 基于伯努利滤波器算法的弱小目标检测方法和装置 |
CN117214857A (zh) * | 2023-11-09 | 2023-12-12 | 中国人民解放军海军航空大学 | 一种三维场景下高斯多假设多扩展目标的跟踪方法 |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102043150A (zh) * | 2010-12-06 | 2011-05-04 | 电子科技大学 | 用于微弱目标检测的改进型粒子滤波检测前跟踪方法 |
CN102621543A (zh) * | 2012-04-02 | 2012-08-01 | 中国人民解放军海军航空工程学院 | 基于粒子滤波和航迹管理的微弱目标检测前跟踪方法 |
CN102621542A (zh) * | 2012-04-02 | 2012-08-01 | 中国人民解放军海军航空工程学院 | 基于多模粒子滤波和数据关联的机动微弱目标检测前跟踪方法 |
CN104076355A (zh) * | 2014-07-04 | 2014-10-01 | 西安电子科技大学 | 基于动态规划的强杂波环境中弱小目标检测前跟踪方法 |
CN106054173A (zh) * | 2016-07-27 | 2016-10-26 | 中国人民解放军海军航空工程学院 | 基于递推Hough变换的微弱多目标检测前跟踪方法 |
CN109991597A (zh) * | 2019-04-04 | 2019-07-09 | 中国人民解放军国防科技大学 | 一种面向弱扩展目标的检测前跟踪方法 |
JP2020008351A (ja) * | 2018-07-04 | 2020-01-16 | 株式会社デンソーアイティーラボラトリ | トラキング装置、方法及びプログラム |
Family Cites Families (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR2378318A1 (fr) * | 1977-01-21 | 1978-08-18 | Thomson Csf | Systeme de poursuite d'une cible mobile |
JP2003524831A (ja) * | 1999-11-22 | 2003-08-19 | イッサム・リサーチ・ディベロップメント・カンパニー・オブ・ザ・ヘブルー・ユニバーシティ・オブ・エルサレム | 組み合わせ空間を探索するためのシステムおよび方法 |
US7719461B1 (en) * | 2008-08-05 | 2010-05-18 | Lockheed Martin Corporation | Track fusion by optimal reduced state estimation in multi-sensor environment with limited-bandwidth communication path |
CN103472445B (zh) * | 2013-09-18 | 2015-06-17 | 电子科技大学 | 一种针对多目标场景的检测跟踪一体化方法 |
CN104101875B (zh) * | 2014-07-11 | 2017-07-11 | 西北工业大学 | 一种空管雷达多目标跟踪方法 |
CN106533451B (zh) * | 2016-11-17 | 2019-06-11 | 中国科学技术大学 | 一种块稀疏信号恢复的迭代终止条件设置方法 |
WO2018119912A1 (zh) * | 2016-12-29 | 2018-07-05 | 深圳大学 | 基于并行模糊高斯和粒子滤波的目标跟踪方法及装置 |
CN108919255A (zh) * | 2018-08-03 | 2018-11-30 | 谭顺成 | 基于cs-pf的高重频雷达弱小目标检测跟踪方法 |
-
2020
- 2020-12-01 CN CN202011383508.6A patent/CN112649802B/zh active Active
-
2021
- 2021-01-20 WO PCT/CN2021/072857 patent/WO2022116375A1/zh active Application Filing
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102043150A (zh) * | 2010-12-06 | 2011-05-04 | 电子科技大学 | 用于微弱目标检测的改进型粒子滤波检测前跟踪方法 |
CN102621543A (zh) * | 2012-04-02 | 2012-08-01 | 中国人民解放军海军航空工程学院 | 基于粒子滤波和航迹管理的微弱目标检测前跟踪方法 |
CN102621542A (zh) * | 2012-04-02 | 2012-08-01 | 中国人民解放军海军航空工程学院 | 基于多模粒子滤波和数据关联的机动微弱目标检测前跟踪方法 |
CN104076355A (zh) * | 2014-07-04 | 2014-10-01 | 西安电子科技大学 | 基于动态规划的强杂波环境中弱小目标检测前跟踪方法 |
CN106054173A (zh) * | 2016-07-27 | 2016-10-26 | 中国人民解放军海军航空工程学院 | 基于递推Hough变换的微弱多目标检测前跟踪方法 |
JP2020008351A (ja) * | 2018-07-04 | 2020-01-16 | 株式会社デンソーアイティーラボラトリ | トラキング装置、方法及びプログラム |
CN109991597A (zh) * | 2019-04-04 | 2019-07-09 | 中国人民解放军国防科技大学 | 一种面向弱扩展目标的检测前跟踪方法 |
Non-Patent Citations (2)
Title |
---|
LI, CUIYUN ET AL.: "Track-before-detect Using Gaussian Particle Probability Hypothesis Density", SYSTEMS ENGINEERING AND ELECTRONICS, vol. 37, no. 4, 30 April 2015 (2015-04-30), pages 740 - 745, XP055937198, ISSN: 1001-506X * |
TAN, SHUN-CHENG ET AL.: "TS-MMPF Based Maneuvering Weak Target TBD Algorithm", FIRE CONTROL & COMMAND CONTROL, vol. 44, no. 3, 31 March 2019 (2019-03-31), pages 38 - 43, XP055937195, ISSN: 1002-0640 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115099343A (zh) * | 2022-06-24 | 2022-09-23 | 江南大学 | 一种有限视野下分布式标签多伯努利融合跟踪方法 |
CN115099343B (zh) * | 2022-06-24 | 2024-03-29 | 江南大学 | 一种有限视野下分布式标签多伯努利融合跟踪方法 |
CN115407299A (zh) * | 2022-09-15 | 2022-11-29 | 中国人民解放军国防科技大学 | 基于伯努利滤波器算法的弱小目标检测方法和装置 |
CN117214857A (zh) * | 2023-11-09 | 2023-12-12 | 中国人民解放军海军航空大学 | 一种三维场景下高斯多假设多扩展目标的跟踪方法 |
CN117214857B (zh) * | 2023-11-09 | 2024-02-02 | 中国人民解放军海军航空大学 | 一种三维场景下高斯多假设多扩展目标的跟踪方法 |
Also Published As
Publication number | Publication date |
---|---|
CN112649802B (zh) | 2022-03-22 |
CN112649802A (zh) | 2021-04-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2022116375A1 (zh) | 高分辨传感器弱小多目标检测前跟踪方法 | |
CN108009628B (zh) | 一种基于生成对抗网络的异常检测方法 | |
CN107516321B (zh) | 一种视频多目标跟踪方法及装置 | |
CN104965199B (zh) | 雷达视频运动目标特征融合判定方法 | |
WO2012009947A1 (zh) | 特征迹驱动的多目标跟踪装置及方法 | |
CN106405537B (zh) | 基于位置信息和多普勒信息的雷达航迹起始方法 | |
CN110555868A (zh) | 一种复杂地面背景下运动小目标检测方法 | |
CN108645412B (zh) | 一种多传感器自适应航迹起始方法 | |
CN106291498B (zh) | 一种基于粒子滤波的检测跟踪联合优化方法 | |
CN107436434B (zh) | 基于双向多普勒估计的航迹起始方法 | |
CN111767847A (zh) | 一种集成目标检测和关联的行人多目标跟踪方法 | |
CN113628245B (zh) | 多目标跟踪方法、装置、电子设备和存储介质 | |
CN111160212A (zh) | 一种基于YOLOv3-Tiny改进的跟踪学习检测***及方法 | |
CN115034324B (zh) | 一种多传感器融合感知效能增强方法 | |
CN111161325A (zh) | 基于卡尔曼滤波与lstm的三维多目标跟踪方法 | |
CN109239702B (zh) | 一种基于目标状态集的机场低空飞鸟数量统计方法 | |
Han et al. | A method based on multi-convolution layers joint and generative adversarial networks for vehicle detection | |
CN113269811A (zh) | 数据融合方法、装置及电子设备 | |
CN116778290A (zh) | 一种基于深度学习算法的雷达视觉数据关联方法 | |
CN111210458B (zh) | 一种基于预检测置信度的运动目标检测前跟踪方法 | |
Li et al. | Fish trajectory extraction based on object detection | |
Guo et al. | Enhanced camera-based individual pig detection and tracking for smart pig farms | |
CN109389543B (zh) | 公交运营数据统计方法、***、计算设备及存储介质 | |
CN115657008A (zh) | 一种机载太赫兹雷达多目标跟踪方法和装置 | |
CN115220002B (zh) | 一种固定单站的多目标数据关联跟踪方法和相关装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21899419 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 21899419 Country of ref document: EP Kind code of ref document: A1 |