WO2019012575A1 - Dispositif de suivi de cible - Google Patents

Dispositif de suivi de cible Download PDF

Info

Publication number
WO2019012575A1
WO2019012575A1 PCT/JP2017/025118 JP2017025118W WO2019012575A1 WO 2019012575 A1 WO2019012575 A1 WO 2019012575A1 JP 2017025118 W JP2017025118 W JP 2017025118W WO 2019012575 A1 WO2019012575 A1 WO 2019012575A1
Authority
WO
WIPO (PCT)
Prior art keywords
target
probability
motion specification
current time
motion
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.)
Ceased
Application number
PCT/JP2017/025118
Other languages
English (en)
Japanese (ja)
Inventor
響介 小西
小幡　康
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Mitsubishi Electric Corp
Original Assignee
Mitsubishi Electric Corp
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Mitsubishi Electric Corp filed Critical Mitsubishi Electric Corp
Priority to PCT/JP2017/025118 priority Critical patent/WO2019012575A1/fr
Priority to JP2019529328A priority patent/JP6570800B2/ja
Publication of WO2019012575A1 publication Critical patent/WO2019012575A1/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

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
    • G01S13/66—Radar-tracking systems; Analogous systems

Definitions

  • the present invention relates to a target tracking device that estimates the number of targets and the motion specification of each target from the reception signals of sensors such as radars and cameras for observing targets such as aircraft and aircraft.
  • the sensor reception signal refers to the observation information of the target acquired by the sensor, and is, for example, the reflected radio wave intensity from the target in the radar, and also the light quantity in each pixel in the infrared camera, for example.
  • a target tracking device for example, as a technology assuming that one target is separated into two targets, a plurality of parallel tracking algorithms in which the number of targets to be assumed and the motion model to be assumed are different There is also a technique of mixing the results of tracking algorithms obtained from the time series of sensor reception signals to estimate the number of targets and motion specifications.
  • Non-Patent Document 1 has a problem that the amount of calculation becomes large by parallelly executing a model assuming before separation and a model assuming after separation.
  • the technique disclosed in Non-Patent Document 1 when one target is first split into two and thereafter two are maintained, at least “maintain target number 1” “target number 1 to 2” As in the case of “increase”, “maintain target number 2”, etc., it is necessary to execute in parallel the tracking algorithm of the same number as the target number is maintained and increased / decreased. As a result, there has been a problem that as the fluctuation of the target number becomes more complicated, the amount of calculation is combinedly increased.
  • An object of the present invention is to provide a target tracking device capable of estimating the motion specification of a target with a small amount of calculation.
  • the target tracking device is based on the motion specification probability distribution of the first target at the past time and the target detection signal information which is observation information of the target, the motion specification probability distribution of the first target at the current time.
  • the motion specification probability distribution of the second target at the current time is predicted The first target is observed at the current time based on the second target motion specification prediction unit, the probability that the first target of the past time is present in the observation area, and the probability that the second target is present in the observation area
  • An existence probability prediction unit that calculates a prediction value of a probability existing in a region and a prediction value of a probability that a second target exists in an observation region, a first target motion specification prediction unit, and a second target motion specification prediction First target and second target of the current time based on the output of the division unit, the existence probability prediction unit, and the target detection signal information Based on the output of the update processor and the update processor that calculates
  • the target tracking device uses the motion specification probability distribution of the first target at the current time based on the motion specification probability distribution of the first target at the past time and the target detection signal information that is observation information of the target.
  • the motion specification probability distribution of the second target at the current time is predicted based on the motion specification probability distribution of the second target which is different from the first target at the past time, and the prediction results of the current target are used to
  • the motion specification probability distribution of the first and second targets at time, the probability that the first target is present in the observation area at the current time, and the probability that the second target is present in the observation area are calculated. It is a thing. This makes it possible to estimate the number of targets and the motion specifications of each target with a small amount of calculation.
  • the target tracking device is targeted at the case where one target appearing at an unknown time separates and emits another target at a certain time. It shall apply.
  • the target on the separating side will be referred to as the "first target”
  • the target on the separated side will be referred to as the "second target”. That is, the case where the target tracking device is applied to the case where the first target appears at a certain time and the first target separates and emits the second target at a certain time will be described.
  • FIG. 1 is a configuration diagram of a target tracking device and its peripheral devices according to Embodiment 1 of the present invention.
  • the target tracking device 100 receives an input from the sensor 200 and sends an output to the display 300.
  • the target tracking device 100 includes a first target motion specification prediction unit 1, a second target motion specification prediction unit 2, an existing probability prediction unit 3, an update processing unit 4, and an estimated value calculation unit 5.
  • time frame the division of the time that the sensor 200 inputs to the target tracking device 100 will be described as “time frame”, and the latest time frame in the process will be “current time frame”, and the time one time frame past from the current time frame
  • the frame is referred to as a "previous time frame”.
  • FIG. 2 shows observation conditions to which the target tracking device 100 according to the first embodiment is applied.
  • the horizontal axis represents the observation time
  • the vertical axis represents the position in the sensor observation area.
  • the detection signal 201 in FIG. 2 represents “a signal detected by the sensor 200 at each time frame”, and is data transmitted from the sensor 200 to the target tracking device 100.
  • the detection signal 201 can be obtained by thresholding the signal strength of each received beam, such as the signal strength exceeding the threshold and the position of the received beam, Doppler velocity, etc. It is.
  • the detection signal 201 is the light amount exceeding the threshold, the position of the pixel, or the like obtained by performing the threshold process on the light amount data of each pixel.
  • the position component of the detection signal 201 output from the sensor 200 is drawn at each observation time.
  • the detection signal 201 may be a reception signal of a sensor that does not pass threshold processing. That is, as a detection signal when the threshold value is set to the minimum value, signal intensity data of all received beams may be input to the target tracking device 100 in the case of radar, or light intensity data of all pixels in the case of a camera. May be input to the target tracking device 100.
  • Loss detection 202 in FIG. 2 represents “an event that was not observed as a detection signal 201 despite the presence of a target”. For example, due to the small size of the target, a loss detection 202 occurs if the strength of the received signal from the target falls below the threshold for detection. The target tracking device 100 presupposes that such a loss detection 202 can occur.
  • the false detection 203 in FIG. 2 is “an unnecessary detection signal caused by something other than the target”. For example, when a received signal derived from topography, noise inside the device, or the like exceeds a detection threshold, false detection 203 occurs. The target tracking device 100 presupposes that such false detection 203 can occur.
  • the first target track 204 in FIG. 2 represents the trajectory of the target separating the second target.
  • the target tracking device 100 assumes that the first target track 204 appears at an unknown position at an unknown time, moves at an unknown speed, separates the second target at an unknown time, and disappears at an unknown time. I assume.
  • the second target track 205 in FIG. 2 represents the trajectory of the separated target.
  • the target tracking device 100 assumes that the second target track 205 is separated from the first target track at an unknown time, travels at an unknown speed, and disappears at an unknown time.
  • at most one detection signal 201 derived from one target is assumed. Further, separation is defined as that the first target and the second target are output from the sensor 200 as different detection signals 201.
  • the sensor 200 is configured of an apparatus for observing the position of a target.
  • the sensor 200 includes existing devices such as a pulse transmission device, a transceiver, an antenna, a receiver, and a signal processing device.
  • the sensor 200 includes an existing device such as a light collecting device, a light receiving device, and an image processing device.
  • the sensor 200 is configured to send a detection signal 201 to the first target motion specification prediction unit 1 and the update processing unit 4.
  • the first target exercise specification prediction unit 1 receives the detection signal 201 which is target detection signal information from the sensor 200, and from the update processing unit 4, the exercise specification probability of the first target in the previous time frame (past time)
  • the distribution 4 a is received, and the predicted value 1 a of the movement specification probability distribution of the first target in the current time frame (current time) is sent to the update processing unit 4.
  • the motion specification probability distribution of the first target is a distribution that represents the probability that the motion specification of the first target becomes a certain value, and, for example, the motions of the position and velocity in the three-dimensional space are estimated.
  • the motion specification probability distribution is a probability density distribution in a total of six-dimensional space of three-dimensional position and three-dimensional velocity.
  • the second target motion specification prediction unit 2 receives from the update processing unit 4 the motion specification probability distribution 4a of the first target in the previous time frame and the motion specification probability distribution 4b of the second target in the previous time frame, The predicted value 2 a of the target motion specification probability distribution is sent to the update processing unit 4.
  • the motion specification probability distribution of the second target is a distribution that represents the probability that the motion specification of the second target becomes a certain value.
  • the existence probability prediction unit 3 receives the existence probability 4c of the first target in the previous time frame and the existence probability 4d of the second target in the previous time frame from the update processing unit 4, and generates the predicted value 3a of the existence probability of the first target , The predicted value 3b of the existence probability of the second target is sent to the update processing unit 4.
  • the “presence probability” is a value of 0 or more and 1 or less that represents the magnitude of the possibility that the target is within the observable range.
  • the update processing unit 4 includes a detection signal 201 from the sensor 200, a predicted value 1a of the motion parameter probability distribution of the first target from the first target motion parameter predicting unit 1, and a second target motion parameter predicting unit 2
  • the predicted value 2a of the motion specification probability distribution of the second target from the second, the predicted value 3a of the existing probability of the first target from the existing probability predicting unit 3, and the predicted value 3b of the existing probability of the second target are received.
  • the update processing unit 4 determines that the first target's motion specification probability distribution 4a in the current time frame, the second target's motion specification probability distribution 4b in the current time frame, and the existence of the first target in the current time frame The probability 4 c and the existence probability 4 d of the second target in the current time frame are calculated, and the calculation result is sent to the estimated value calculation unit 5.
  • the update processing unit 4 sets the first target motion specification prediction unit 1 and the second target motion specification prediction unit 1 as the first target's motion specification probability of the first target in the current time frame. It sends to the target motion specification prediction unit 2.
  • the update processing unit 4 sends the second target motion specification probability distribution 4b of the second target in the current time frame to the second target motion specification prediction unit 2 as the motion specification probability of the second target in the previous time frame.
  • the update processing unit 4 also determines the existence probability 4c of the first target in the current time frame and the existence probability 4d of the second target in the current time frame, the existence probability of the first target in the previous time frame, and the second target in the previous time frame. It sends to the existing probability predicting unit 3 as the existing probability of.
  • the estimated value calculation unit 5 calculates the first target motion specification probability distribution 4 a in the current time frame, the second target motion specification probability distribution 4 b in the current time frame, and One target presence probability 4c and a second target presence probability 4d in the current time frame are received. Then, the estimated value calculation unit 5 obtains the motion specification estimated value 5a in the current time frame, and sends this to the display 300.
  • “motion specification estimates” are motion specifications of each target estimated to exist in the observation area. For example, in the case of estimating position and velocity as movement parameters, if it is estimated that only the first target exists in the estimated value calculation unit 5, the movement parameter estimation value is the position and velocity of the first target.
  • the motion specification estimated values are the position and the velocity of the first target and the position and the velocity of the second target. If it is estimated by the estimated value calculation unit 5 that there is no target, the motion specification estimated value is data representing no value.
  • the display 300 comprises an existing device for displaying the motion specification estimated value 5a sent from the estimated value calculation unit 5 to the user. For example, it is a display etc. which display an image. Further, an apparatus for recording the exercise specification estimation value 5a as an image for displaying to the user, and a processing apparatus for combining with other information are also examples of the display 300.
  • the display 300 receives the motion specification estimated value 5a in the current time frame from the estimated value calculation unit 5, and displays information such as the motion specification of each target to the user.
  • FIG. 3 is a diagram showing an example of a hardware configuration of the target tracking device 100 according to the first embodiment.
  • the illustrated target tracking device 100 includes an arithmetic device 101, a recording device 102, an input interface 103, an output interface 104, and a bus 105.
  • the arithmetic device 101 is configured of a processor such as a central processing unit (CPU) or a graphics processing unit (GPU).
  • the recording device 102 is a storage device such as a hard disk drive (HDD) or a solid state drive (SSD), a storage device such as a dynamic random access memory (DRAM), or a flash memory.
  • the input interface 103 is a USB (Universal Serial Bus), Ethernet (registered trademark), or the like.
  • the output interface 104 is, for example, DVI (Digital Visual Interface: registered trademark), HDMI (High-Definition Multimedia Interface: registered trademark), USB (Universal Serial Bus), Ethernet (registered trademark), or the like.
  • the bus 105 is a communication path for connecting the arithmetic device 101 to the output interface 104 to each other.
  • the programs corresponding to the respective functions are stored in the recording device 102, and the programs corresponding to the processing are read out by the arithmetic device 101 and executed. It is realized by being done.
  • the arithmetic device 101 writes, reads, and deletes input / output values and processing intermediate data in the recording device 102 while the program is being executed.
  • the detection signal 201 from the sensor 200 is taken into the target tracking device 100 via the input interface 103. That is, the first target motion specification prediction unit 1 and the update processing unit 4 acquire the detection signal 201 via the input interface 103.
  • the output from the estimated value calculation unit 5 is sent to the display 300 via the output interface 104. That is, the motion specification estimated value from the estimated value calculation unit 5 is transmitted to the display 300 via the output interface 104.
  • FIG. 4 shows operations of the first target motion specification prediction unit 1, the second target motion specification prediction unit 2, the existence probability prediction unit 3, the update processing unit 4, and the estimated value calculation unit 5 in the current time frame. It is a flowchart.
  • FIG. 5 is a flowchart showing the operation of the update processing unit 4 in the current time frame, and shows the operation of step ST4 of FIG. 4 in more detail.
  • the premise in the following operation explanation will be shown below.
  • the coordinate system of the three-dimensional position space is orthogonal XYZ coordinate axes, and the velocity component representing the target velocity is also each direction of the XYZ axes.
  • the motion specification in the target tracking device 100 is not limited to this example, and for example, for a target moving in a two-dimensional plane, the position in the two-dimensional position space and the velocity in the two-dimensional velocity space It may be assumed that Also, for example, when the acceleration of the target is also estimated, the acceleration component may be included in the movement specification. Also, polar coordinates or the like may be used instead of the orthogonal coordinate system. In addition, in the following, the target predicts movement specifications based on the premise of constant-speed linear movement. Note that the premise of the target tracking device 100 is not limited to this example, and instead of constant velocity straight movement, motion specifications may be predicted on the assumption of, for example, constant acceleration straight movement, constant angular velocity swing movement, and the like.
  • the detection signal 201 input from the sensor 200 to the target tracking device 100 is the position of the detected signal and the intensity of the detected signal.
  • the premise of the input in the target tracking device 100 is not limited to this.
  • the sensor 200 includes a radar device
  • the Doppler component of the detected signal may be used as the input of the target tracking device 100.
  • the sensor 200 includes a camera
  • the light quantity of the detected signal for each color system may be used as the input of the target tracking device 100.
  • the motion parameter probability distribution is approximated as a linear combination of Gaussian distributions (strictly speaking, “approximated as a linear combination of probability density functions of Gaussian distributions”, but in the following, for simplification of the notation “Probability density function of Gaussian distribution” is described as “Gaussian distribution”.
  • the method of approximating the motion parameter probability distribution in the target tracking device 100 is not limited to this, and, for example, B. Vo, B. Vo, D.
  • the number of the previous time frame is k-1 and represented as "time k-1.”
  • x (1) , x (2) and x (3) represent the X-axis component, Y-axis component and Z-axis component of the position
  • x (4) , x (5) and x (6) are the velocity Represents an X-axis direction component, a Y-axis direction component, and a Z-axis direction component of
  • T in the upper right subscript represents transposition of a matrix.
  • a vector representing such motion specification will be referred to as a "state vector".
  • the first target motion specification prediction unit 1 estimates the predicted value of the first target motion specification probability distribution at time k from the detection signal 201 and the first target motion specification probability distribution at time k-1. calculate.
  • the Gaussian distribution that constitutes the predicted value of the first target motion specification probability distribution at time k is classified into two types. One is a Gaussian distribution representing the possibility that "the first target has appeared in the past from time k", and is calculated from the first target motion specification probability distribution at time k-1. The other is a Gaussian distribution representing the possibility that "the first target appears at time k", and is calculated from the detection signal 201 in a time frame in the past from time k.
  • the Gaussian distribution representing the possibility that "the first target has appeared in the past from time k" is suffixed with "S”
  • a Gaussian distribution that represents the possibility of "the first target appeared at time k" is It represents.
  • k-1 Gaussian distributions representing the possibility that "the first target has appeared in the past from time k", and each weighting coefficient, Gaussian distribution average value, Gaussian distribution
  • the covariance matrix is calculated from the following equation.
  • j 1, 2,..., J k ⁇ 1
  • ⁇ k is a matrix that causes the state vector to transition from time k-1 to time k based on a model that the target moves at a constant speed and rectilinearly, It is.
  • ⁇ k represents an elapsed time between time k-1 and time k.
  • Q k is a matrix that represents the ambiguity of the model that the target moves at a constant linear motion, and is a parameter of 6 rows and 6 columns.
  • the k-2 Gaussian distributions, and their respective weighting factors, Gaussian distribution mean values, and Gaussian distribution covariance matrix are calculated from the following equations.
  • w init is a parameter representing the initial value of the weighting coefficient
  • ⁇ k-1 is an elapsed time between time k-2 and time k-1
  • P init is the initial value of the Gaussian distribution covariance matrix 6 It is a parameter of row 6 column.
  • Gaussian distributions ie Is the predicted value of the first target motion specification probability distribution at time k (the symbol ⁇ represents a union).
  • Each weighting coefficient in the predicted value of the first target motion specification probability distribution is standardized at the end of this step so as to satisfy the following equation.
  • the second target motion specification prediction unit 2 calculates the second target motion specification probability distribution at time k-1 and the second target motion specification probability distribution at time k-1
  • the predicted value of the target motion specification probability distribution is calculated.
  • the Gaussian distribution that constitutes the predicted value of the second target motion specification probability distribution at time k is classified into two types. One is a Gaussian distribution representing the possibility of “the second target has been separated in the past from time k”, and is calculated from the second target motion specification probability distribution at time k ⁇ 1. The other is a Gaussian distribution representing the possibility of “the second target is separated at time k”, which is calculated from the first target motion specification probability distribution at time k ⁇ 1.
  • the Gaussian distribution representing the possibility that “the second target has been separated in the past from time k” is given the suffix "S"
  • k-1 (2) Gaussian distributions representing the possibility that “the second target has been separated in the past from time k”, and each weighting coefficient, Gaussian distribution average value, Gaussian
  • the distribution covariance matrix is calculated from the following equation.
  • j 1, 2,..., J k ⁇ 1
  • k k and Q k are the same as step ST1.
  • k-1 (1) Gaussian distributions representing the possibility of “the second target is separated at time k”, and the respective weighting coefficients, the Gaussian distribution average value, the Gaussian
  • the distribution covariance matrix is calculated from the following equation.
  • j 1, 2,..., J k ⁇ 1
  • w init and P init are the same as described above.
  • ⁇ A is a parameter of 6 rows and 6 columns giving “the moving direction immediately after the separation of the second target”, for example It is.
  • ⁇ 1 , ⁇ 2 and ⁇ 3 are parameters representing an X-axis rotation angle, a Y-axis rotation angle and a Z-axis rotation angle, respectively, from the moving direction of the first target.
  • ⁇ B is a 6 ⁇ 6 parameter giving “the speed immediately after the separation of the second target”, and, for example, It is.
  • ⁇ is a parameter representing the ratio of the second target velocity to the first target velocity.
  • a plurality of parameters related to the velocity component immediately after separation of the second target, for example, ⁇ 1 to ⁇ 3 and ⁇ may be set. In that case, J k ⁇ 1
  • the Gaussian distribution representing the possibility of “the second target has been separated in the past from time k” calculated by the above and the Gaussian distribution representing the possibility of “the second target being separated in time k” are combined Total J k
  • the weighting coefficient in the predicted value of the second target motion specification probability distribution is standardized at the end of this step so as to satisfy the following equation.
  • step ST3 the existing probability predicting unit 3 predicts the existing probability of the first target at time k from the existing probability of the first target at time k-1 and the existing probability of the second target at time k-1.
  • the predicted value of the existence probability of the second target at time k is calculated.
  • FIG. 6 shows a model of a stochastic process used when predicting the existence probability of each target in step ST3.
  • processing for predicting the existence probability of the first target and the second target will be described according to FIG.
  • FIG. 6 shows a model that transits between observation time frames with a specific probability between the four states S1 to S4 in which the presence or absence of each target is different.
  • S1 represents "a state in which neither a first goal nor a second goal exists”.
  • S2 represents "a state in which a first target exists and a second target does not exist”.
  • S3 represents "a state in which both the first goal and the second goal exist”.
  • S4 represents "a state in which the first target does not exist and the second target exists”.
  • S0 represents an initial state, and in this example, it indicates that the state starts from S1 in the time frame at the start of observation.
  • arrows between the states indicate paths of transition.
  • the transition path TR12 is a transition from S1 to S2, and represents a transition of "the first target appears and the second target does not appear".
  • the transition path TR22 is a transition from S2 to S2, and represents a transition of "the first target does not disappear and continues to exist, and the second target does not appear.”
  • transition between some states is prohibited.
  • the transition route from S1 to S3, the transition route from S4 to S2, and the transition route from S4 to S3 are also prohibited based on the same idea.
  • step ST4 the update processing unit 4 detects the detection signal 201 at time k, the predicted value of the first target exercise specification probability distribution at time k, and the predicted value of the second target exercise specification probability distribution at time k And the predicted value of the existence probability of the first target at time k and the prediction value of the existence probability of the second target at time k, the motion specification probability distribution of the first target at time k, and the second target at time k The motion specification probability distribution of the first target, the existence probability of the first target at the time k, and the existence probability of the second target at the time k are calculated.
  • the process in step ST4 is the same as the process disclosed in the document 1.
  • FIG. 5 shows the process flow in step ST4 in more detail.
  • the processes in steps ST41 to ST44 in FIG. 5 will be described below.
  • step ST41 the predicted value of the existence probability of the nth target And the confidence level for each subset of goal numbers
  • the subset reliability in step ST41 is calculated from the following equation. Where on the right side of the equation
  • step ST42 the detection signal at time k is And the subset reliability And the predicted value of the motion specification probability distribution of the nth target And the degree of confidence for each mapping And the motion specification probability distribution of the nth target for each mapping.
  • the prime (“'") attached to the symbol of the reliability and the weighting factor is used as a symbol indicating that it is a value before being standardized. The method of standardization will be described later in step ST43.
  • step ST42 The value calculated by step ST42 is obtained from the following equation.
  • R in the equation (41) is a 6 ⁇ 6 parameter representing an error covariance with respect to the position component of the detection signal.
  • pk and DT in equation (39) are parameters representing the probability of detection without loss of the target at time k
  • ⁇ k and FLS in equation (39) are per unit volume at time k It is a parameter that represents the number of false detections.
  • ⁇ (z k ( ⁇ (n), 4) ) in the equation (39) is the value of “intensity component z k ( ⁇ (n), 4) ” for the intensity component z k ( ⁇ (n), 4 ) of the detection signal.
  • detection signal is the target probability " ⁇ " intensity components z k of) (theta (n), is the likelihood ratio representing the probability "is erroneously detected detection signal of 4).
  • the specific value of this ⁇ (z k ( ⁇ (n), 4) ) differs depending on the type of sensor, and for example, if the detection signal is a value acquired by a radar device, D. Lerro, Y. Bar-Shalomo , "Automated Tracking with Target Amplitude Information," American Control Conference 1990, pp. 2875-2880, San Diego, May 1990. (hereinafter referred to as "Document 2").
  • step ST43 the degree of reliability for each mapping And the weighting factor in the motion specification probability distribution of the nth target for each mapping Standardize each When Calculate The normalization in step ST43 is performed by the following equation.
  • step ST44 the reliability for each standardized mapping is obtained. And the motion specification probability distribution of the nth target for each mapping whose weighting factor is normalized And the probability of the existence of the nth target at time k And the motion specification probability distribution of the nth target at time k Are calculated, and the process of step ST4 is finished.
  • step ST44 The value calculated by step ST44 is obtained from the following equation. here, I assume. Also, on the right side of equation (46) In addition, since the number of Gaussian distributions representing the motion specification probability distribution monotonously increases each time step ST4 is executed, processing for simplifying the shape of the motion specification probability distribution of each target is added after step ST4. It is also good. Specifically, B. Vo, W. Ma, "The Gaussian mixture probability hypothesis density filter," IEEE Trans. Signal Process., Vol. 54, no. 11, pp. 4091-4104, Nov. 2006. 3) deletion of a Gaussian distribution whose weighting factor is smaller than a predetermined value, and processing of combining Gaussian distributions having similar average values into one Gaussian distribution.
  • step ST5 the estimated value calculation unit 5 determines whether the existence probability of the first target at time k exceeds the existence probability threshold. That is, it is determined whether the following equation is true or false.
  • r Th (1) is a parameter having a meaning that “If the probability of existence of the first target is larger than this value, the first target is presumed to exist”. If the above determination is true, step ST6 is executed, and if false, the process proceeds to step ST7.
  • step ST6 the estimated value calculation unit 5 estimates the movement specification of the first target at time k from the movement specification probability distribution of the first target at time k.
  • iEx represents "the number of the Gaussian distribution having the largest weighting coefficient among the Gaussian distributions constituting the first target motion specification probability distribution". That is, when expressed by a formula, It is.
  • step ST7 the estimated value calculation unit 5 determines whether the existence probability of the second target at time k exceeds the existence probability threshold. That is, it is determined whether the following equation is true or false.
  • r Th (2) is a parameter having a meaning of “estimates that the second target is present if the probability of the presence of the second target is larger than this value”. If the above determination is true, step ST8 is executed, and if false, the operation of the target tracking device 100 at time k is ended.
  • step ST8 the estimated value calculation unit 5 estimates the movement specification of the second target at time k from the movement specification probability distribution of the second target at time k.
  • i Ex represents “the number of the Gaussian distribution having the largest weighting coefficient among the Gaussian distributions constituting the movement specification probability distribution of the second target”. That is, when expressed by a formula, It is.
  • the following effects can be obtained.
  • the second target motion specification prediction unit 2 in addition to the predicted values of the motion specification probability distribution of the second target at the current time, other than the motion specification probability distribution of the second target at the previous time, It was configured to calculate also from the movement specification probability distribution of the first target at the previous time.
  • the predicted value based on the law of inertia that is, when the second target is separated from the first target, motion parameters such as the position and velocity of the second target immediately after separation are close to the first target.
  • the predicted value of the motion specification probability distribution of the second target conforming to the physical law is obtained.
  • the accuracy of the second target motion specification estimated value calculated by the update processing unit 4 and the estimated value calculation unit 5 based on the predicted value is, for example, as described in Document 1 and DB Reid, “An Algorithm for Tracking Multiple Targets,” Motion parameter prediction values for each target as disclosed in IEEE Trans. Automatic Control, vol. 24, no. 6, pp. 843-854, Dec., 1979. (hereinafter referred to as reference 4)
  • reference 4 Compared with the prior art that estimates without using the motion specification prediction value of the target of the target, it is an estimated value obtained from a prediction value based on a model close to real physical law, so the error with the true motion specification is small . That is, even if the observation time from separation is short, or even if the number of times of detection of the second target is small due to a loss detection, the motions of the second target with higher accuracy than the techniques disclosed in Document 1 and Document 4 You can get the source.
  • the existence probability of the second target is used when calculating the prediction value of the existence probability of the second target. It was configured to use the probability of existence. According to this configuration, the predicted value of the existence probability of each target is obtained based on the premise that the second target appears after the first target appears.
  • the number of motion specifications of each target estimated by the update processing unit 4 and the estimated value calculation unit 5 based on the predicted values is, for example, each target as disclosed in the documents 1 and 4 It is obtained from the predicted value based on a model that simulates the order of actual target appearance, compared to the prior art in which it is presumed that the stochastic process of appearance and disappearance is independent of the appearance and appearance of another target. Since the estimated value is calculated, the error from the actual target number is small.
  • the first target motion specification prediction unit 1 and the second target motion specification prediction unit 2 calculate predicted values of the motion specification probability density distribution for each target, and the presence probability prediction unit 3
  • the predicted value of the existence probability is calculated for each target
  • the update processing unit 4 calculates the exercise specification probability distribution and the existence probability for each target
  • the estimated value calculation unit 5 calculates the exercise specification estimated value for each target Configured to calculate
  • the "first target exercise specification” that maintains the previous exercise and the "second target exercise specification” separated from the first goal are obtained separately.
  • This effect is particularly important in prioritizing goals after separation. For example, when applied to a condition in which another target is emitted from a target moving in a parabolic trajectory, a first target moving in a parabolic trajectory as in the case before separation, and a second target emitted with acceleration at the time of separation Can distinguish. Therefore, for example, when the interest with respect to the first target maintaining the parabolic orbit is high and the first target needs to be additionally observed, it is possible to avoid the situation where the second target is erroneously additionally observed.
  • the presence probability predicting unit 3 is configured to predict the presence probability by the equations (29) and (30) based on the model of FIG. According to this configuration, the target number is maintained and fluctuates as disclosed in Non-Patent Document 1 under observation conditions in which one target is divided into two and the target number fluctuates between 0 and 2. (29) and (30), each target can be calculated with a small amount of operation as compared with the prior art in which it is necessary to execute in parallel the tracking algorithm of the same number as the number of transition paths TR11 to TR44 in FIG. Motion specifications can be estimated.
  • the second target motion specification prediction unit 2 predicts the predicted value of the second specification's motion specification probability distribution at the current time, the motion value probability distribution of the first target at the previous time, and It was configured to calculate from the motion specification probability distribution of the second target at time. With this configuration, even when the observation time is short or the movement specification of the first target is not narrowed down due to the detection of the loss of the first target, etc., the movement specification with the possibility of the second target is predicted. 2) It is possible to estimate the presence or absence of a goal. Therefore, for example, as disclosed in Japanese Patent Application Laid-Open No.
  • Patent Document 1 the movement target of the separating target, that is, the first target's movement specification is referred to as "the main track”.
  • the main track the movement target of the separating target, that is, the first target's movement specification.
  • the target tracking device of the first embodiment based on the first target motion specification probability distribution of the past time and the target detection signal information which is observation information of the target,
  • the second target of the current time based on the first target motion specification prediction unit that predicts the motion specification probability distribution of the first target, and the motion specification probability distribution of the second target that is different from the first target of the past time
  • An existence probability prediction unit for calculating a predicted value of the probability that the first target is present in the observation area at the current time and a predicted value of the probability that the second target is present in the observation area; Based on the output of the target unit, the second target motion specification prediction unit, the presence probability prediction unit, and the target detection signal information.
  • Update processing unit that calculates the motion specification probability distribution of the first target and the second target at time, the probability that the first target exists in the observation area at the current time, and the probability that the second target exists in the observation area
  • an estimated value calculation unit that determines whether each target is present in the observation area at the current time based on the output of the update processing unit, and calculates an estimated movement parameter value of the target determined to be present. Therefore, it is possible to estimate the number of targets and the motion specifications of each target with a small amount of calculation.
  • the second target motion specification prediction unit is configured to calculate the motion specification probability distribution of the first target at the past time and the movement specification of the second target at the past time Since the motion specification probability distribution of the second target at the current time is predicted based on the probability distribution, the motion specification of the second target with high accuracy can be obtained.
  • the existence probability prediction unit observes the first target at the current time according to the probability transition model between states regarding the presence or absence of the first target and the presence or absence of the second target. Since the predicted value of the probability existing in the area and the predicted value of the probability that the second target exists in the observation area at the current time are calculated, the motion specification of each target can be estimated with a small amount of calculation. it can.
  • Second Embodiment In the first embodiment, assuming that one target appearing at an unknown time can be divided into two at an unknown time, the motion specification probability distribution and the existence probability of each target are sequentially calculated, I was trying to estimate the movement specification of each target at time. However, depending on the type of goal, one goal may be separated into three or more goals. In such a case, in the configuration of the first embodiment, although there are actually three or more targets, the operation is performed to estimate motion specifications based on the premise that the target number is at most two. There is a problem that at most, only two target movement specifications can be obtained. Therefore, in the second embodiment, on the premise that one target is divided into N maximum targets, the motion specification probability distribution and the existence probability of each target are calculated, and the motion specification of each target is estimated.
  • the maximum target number N is a parameter.
  • first target the target on the separation side
  • FIG. 7 is a block diagram of the target tracking device 100a according to the second embodiment and its peripheral devices.
  • the target tracking device 100 a is configured to receive an input from the sensor 200 and send an output to the display 300.
  • the sensor 200 and the display 300 have the same configuration as in the first embodiment.
  • the target tracking device 100a according to the second embodiment includes a first target motion specification prediction unit 1, second to N target motion specification prediction units 6, an existence probability prediction unit 30, an update processing unit 40, and an estimated value calculation unit. It has 50.
  • the first target motion specification prediction unit 1 is the same as that of the first embodiment, the description here is omitted.
  • the 2nd to Nth target motion specification prediction unit 6 receives from the update processing unit 40 the motion specification probability distribution 40a of the first target in the previous time frame and the motion specification of the second to Nth targets in the previous time frame.
  • the probability distribution 40b is received, and the predicted value 6a of the motion specification probability distribution of the second target to the Nth target is sent to the update processing unit 40.
  • the motion specification probability distribution of the second target to the Nth target represents each motion specification probability distribution of the second target to the Nth target.
  • the existence probability prediction unit 30 receives the existence probability 40c of the first target in the previous time frame and the existence probability 40d of the second target to the Nth target in the previous time frame from the update processing unit 40, and the existence probability of the first target
  • the predicted value 30a and the predicted value 30b of the existence probability of the second target to the Nth target are sent to the update processing unit 40.
  • Presence probability of the second to Nth targets represents the probability of each of the second to Nth targets.
  • the update processing unit 40 detects the detection signal 201 from the sensor 200, the predicted value 1a of the motion parameter probability distribution of the first target from the first target motion parameter predicting unit 1, and the second to N target motion parameter predictions.
  • Predicted value 6a of the motion specification probability distribution of the second target to the Nth target from the part 6 and the predicted value 30a of the existence probability of the first target from the existing probability prediction unit 30 and the existence of the second target to the Nth target The predicted value 30b of the probability is received.
  • the update processing unit 40 generates the motion parameter probability distribution 40a of the first target in the current time frame, the motion parameter probability distribution 40b of the second target to the Nth target in the current time frame, and the The existence probability 40c of one target and the existence probability 40d of the second to Nth targets in the current time frame are calculated, and the calculation result is sent to the estimated value calculation unit 50. Further, the update processing unit 40 sets the first target motion specification prediction unit 1 and the second target motion specification unit 1 as the first target motion specification probability of the first target in the current time frame as the first target motion specification probability. N Send to target motion specification prediction unit 6.
  • the update processing unit 40 sets the motion parameter probability distribution 40b of the second target to the Nth target in the current time frame as the motion parameter probability of the second target to the Nth target in the previous time frame. Send to specification prediction unit 6.
  • the update processing unit 40 determines the existence probability 40c of the first target in the current time frame and the existence probability 40d of the second target to the Nth target in the current time frame, and the existence probability of the first target in the previous time frame and the previous time It is sent to the existing probability predicting unit 30 as the existing probability of the second to Nth targets in the frame.
  • the estimated value calculation unit 50 calculates the first target motion specification probability distribution 40a in the current time frame, the second target to the Nth target motion specification probability distribution 40b in the current time frame, and The existence probability 40c of the first target in the time frame and the existence probability 40d of the second to Nth targets in the current time frame are received, and it is determined whether each target is present in the observation area in the current time frame.
  • the motion specification estimated value 50a of the determined target is calculated.
  • the hardware configuration of the target tracking device is realized by the configuration shown in FIG. That is, the first target motion specification prediction unit 1, the second to N target motion specification prediction units 6, the existence probability prediction unit 30, the update processing unit 40, and the estimated value calculation unit 50 in FIG. 7 correspond to their respective functions. These programs are stored in the recording device 102, and are realized by the arithmetic device 101 reading and executing programs corresponding to the processing. The configuration of each part is the same as that of the first embodiment, and thus the description thereof is omitted.
  • FIG. 9 shows operations of the first target motion specification prediction unit 1, the second to N target motion specification prediction units 6, the existence probability prediction unit 30, the update processing unit 40, and the estimated value calculation unit 50 in the current time frame.
  • the motion specification of the estimation target is the position and velocity in the three-dimensional position space
  • the coordinate system is an orthogonal coordinate system
  • the motion of the target is assumed to be straight at a constant velocity
  • the detection signal 201 is the position of the detected signal.
  • the value be an intensity
  • the motion specification probability distribution be approximated by a linear combination of Gaussian distribution.
  • the definitions of symbols in the following description are the same as in the first embodiment.
  • “n” representing the target number is defined as 1 or more and 2 or less, but in the second embodiment, the range of n is 1 or more and N or less.
  • Step ST1 performed by the first target motion item prediction unit 1 is the same process as step ST1 of FIG. 4 in the first embodiment.
  • step ST10 the 2nd to Nth target motion item prediction portion 6-n inside the 2nd to Nth target motion item prediction portion 6 is at time k-1
  • the predicted value of the nth target motion specification probability distribution at time k is calculated from the first target motion specification probability distribution and the nth target motion specification probability distribution at time k-1.
  • n is an integer of 2 or more and N or less
  • n is set to 2 in step ST9, and n is determined to be n in step ST11.
  • step ST10 is the same as the one in which the input / output of the calculation in step ST2 in the first embodiment is replaced with the “second target” from the “nth target”. That is, in this step ST9, a Gaussian distribution representing the possibility of "the nth target has been separated in the past from time k" Gaussian distribution representing the possibility of “the nth target being separated at time k” Calculate
  • a Gaussian distribution representing the possibility of “the nth target has been separated in the past from time k” is calculated by the following equation.
  • j 1, 2,..., J k ⁇ 1
  • a Gaussian distribution that represents the possibility of “the nth target being separated at time k” is calculated by the following equation. It is.
  • j 1, 2,..., J k ⁇ 1
  • the definitions of k k , Q k , w init , P init , ⁇ A and ⁇ B are the same as in the first embodiment.
  • Gaussian distribution representing the possibility that "the nth target has appeared in the past from time k" calculated by the above formulas (56) to (61), and "the nth target has appeared in the time k" Total of J k
  • the weighting factor in the predicted value of the nth target motion specification probability distribution is assumed to be standardized so as to satisfy the following equation at the end of this step.
  • the existence probability predicting unit 30 determines the existence probability of the first target at time k ⁇ 1 and the existence probability of the second target to the Nth target at time k ⁇ 1.
  • a predicted value of the existence probability of the target and a predicted value of the existence probability of the second to Nth targets at time k are calculated.
  • the predicted value of the existence probability of the nth target is Calculated.
  • p b (1) and p s (1) are the same as in the first embodiment.
  • the probability that the nth target appears between time frames is p b (n)
  • the probability that the nth target disappears between time frames is p s (n) .
  • These probabilities p b (n) and p s (n) may be parameters set in advance or may be values that change with time according to observation conditions.
  • Step ST4 which is the processing subsequent to step ST13, is the same processing as step ST4 of FIG. 4 in the first embodiment except for the maximum number of targets. That is, although the maximum value N of n representing the target number is 2 in the description of the process of step ST4 of the first embodiment, in the second embodiment, this N is a parameter N representing the maximum value of the target number. It is the same as the processing when replacing.
  • Step ST5 is the same process as step ST5 of FIG. 4 in the first embodiment. Further, step ST6 is the same process as step ST6 of FIG. 4 in the first embodiment.
  • the estimated value calculation unit 50 determines whether the existence probability of the nth target at time k exceeds the existence probability threshold value in step ST15. That is, it is determined whether the following equation is true or false.
  • the estimated value calculation unit 50 estimates the movement specification of the nth target at time k from the movement specification probability distribution of the nth target at time k.
  • n is an integer of 2 or more and N or less.
  • iEx represents “the number of the Gaussian distribution having the largest weighting coefficient among the Gaussian distributions constituting the motion parameter probability distribution of the nth target”. That is, when expressed by a formula, It is.
  • the predicted values of the motion specification probability distribution of the second target to the Nth target at the current time in the second to Nth target motion specification prediction unit 6 are the movement specifications of the first target at the previous time. It is configured to calculate from the probability distribution and the motion specification probability distribution of the second target to the Nth target at the previous time.
  • the existing probability predicting unit 30 predicted values of the existing probability of the second target to the Nth target at the current time, the motion specification probability distribution of the first target at the previous time, and the second target to the Nth target at the previous time It is configured to calculate from the existence probability of In this configuration, even if at most N-1 targets can be separated from one target, that is, even if the second to the Nth targets can be separated from the first target, the first target and the second target An estimated value of movement specification determined to be present among the presence and absence of the Nth target and the first and second targets to the Nth target is obtained.
  • This effect is particularly the maximum target number after separation even if an unknown number of alternative targets can be separately emitted from targets such as aircraft appearing at unknown times, or even if the separation of targets may occur an unknown number of times. If even the assumption that the number does not exceed N is valid, the point at which the motion specifications of each target at each time can be estimated is obtained as a great effect.
  • the second to Nth target motion specification prediction unit 6 predicts the predicted values of the motion specification probability distribution of the second to the Nth targets at the current time, the second targets to the Nth at the previous time, and so on. In addition to the motion parameter probability distribution of the target, calculation was also made from the motion parameter probability distribution of the first target at the previous time.
  • the predicted value based on the law of inertia that is, when the second target to the Nth target are separated from the first target, the movement specifications such as the position and the speed of the second to Nth targets immediately after the separation are Predicted values of motion specification probability distributions of the second to Nth targets in accordance with the physical law of being close to the first target are obtained.
  • the estimation accuracy of motion specifications of the second to Nth targets estimated by the update processing unit 40 and the estimated value calculation unit 50 based on the predicted values is, for example, as disclosed in documents 1 and 4.
  • the motion specification prediction value of each target is estimated independently of the motion specification prediction value of another target, it is an estimated value obtained from a prediction value based on a model close to the physical laws of reality,
  • the error with the true movement specification is small. That is, even if the observation time from separation is short, or even if the number of times of detection of the second target is small due to the missing detection 202, the second target to the second with better accuracy compared to the techniques disclosed in Document 1 and Document 4. N You can get the movement specification of the target.
  • the existence probabilities of the second to Nth targets are used, and the second to Nth targets are used.
  • the existence probability of the first target is used. According to this configuration, a predicted value of the existence probability of each goal is obtained based on the premise that the second goal to the Nth goal appear after the first goal appears.
  • the number of motion specifications of each target estimated by the update processing unit 40 and the estimated value calculation unit 50 based on the predicted value is, for example, as disclosed in documents 1 and 4 It is obtained from the predicted value based on a model based on a model that is closer to the order of actual target appearance, compared to the prior art in which the stochastic process in which a target appears and disappears is assumed to be independent of the appearance and another stochastic process of another target. Since the estimated value is calculated, the error with respect to the target number that is truly present is small.
  • the motion specifications of the actually existing target can not be estimated as compared with the techniques disclosed in the documents 1 and 4. It is possible to reduce the frequency of missing out and the frequency of outputting estimates of non-existent targets.
  • the existence probability prediction unit 30 is configured to predict the existence probability by the equations (64) and (65) based on a model obtained by expanding FIG.
  • one target separates up to N-1 additional targets, and the target number as disclosed in Non-Patent Document 1 under observation conditions where the number of targets varies between 0 and N.
  • Equations (64) and (65) make it possible to estimate the motion specification of each target with a small amount of calculation, as compared with the prior art in which the tracking algorithm is executed in parallel with the number in the case where the number is maintained and fluctuates.
  • the predicted values of the motion specification probability distribution of the second target to the Nth target at the current time in the second to Nth target motion specification prediction unit 6 It is configured to calculate from the original probability distribution and the motion specification probability distribution of the second target to the Nth target at the previous time.
  • the movement specification of the separation target that is, the second to Nth targets is disclosed.
  • the movement specification of the separation target that is, the second to Nth targets.
  • the first target motion specification prediction unit that predicts the motion specification probability distribution of the first target, and the nth target (n is an integer of 2 or more and N or less) different from the first target
  • Presence probability prediction unit for calculating the predicted value of the probability existing in the first target motion specification prediction unit and the second Based on the output of the target motion specification prediction unit and the presence probability prediction unit and the target detection signal information, the motion specification probability distribution of the first target
  • the second to Nth target motion specification prediction units are configured to calculate the motion specification probability distribution of the first target at the past time and the motion of the nth target at the past time Since the motion data probability distribution of the nth target at the current time is predicted based on the data probability distribution, the motion data of the nth target with high accuracy can be obtained.
  • the presence probability prediction unit observes the first target at the current time according to the probability transition model between the states regarding the presence or absence of the first target and the presence or absence of the nth target. Since the predicted value of the probability existing in the area and the predicted value of the probability that the nth target exists in the observation area at the current time are calculated, the motion specification of each target can be estimated with a small amount of calculation. it can.
  • the predicted value of the motion specification probability distribution of the first target the motion specification probability distribution of the first target in the past, (2) Calculated from the distribution of probability of movement of target.
  • the combined target is the “first target” and the combined target is the “second target”.
  • a goal whose appearance is early is taken as a first goal
  • a goal whose appearance is late is taken as a second goal.
  • the target tracking device is applied when the first target appears at a certain time, the second target appears at a later time, and the second target is combined with the first target at a later time Do.
  • the second goal may occur even if it disappears without being combined with the first goal.
  • FIG. 10 is a block diagram of a target tracking device 100b according to the third embodiment.
  • the target tracking device 100b is configured to receive an input from the sensor 200 and send an output to the display 300.
  • the target tracking device 100 b includes a first target motion specification prediction unit 10, a second target motion specification prediction unit 20, an existing probability prediction unit 3, an update processing unit 4, and an estimated value calculation unit 5.
  • the configuration other than the first target motion specification prediction unit 10 and the second target motion specification prediction unit 20 is the same as that of the first embodiment shown in FIG. The explanation is omitted.
  • the configuration of the sensor 200 and the display 300 is the same as that of the first embodiment.
  • the detection signal 201 from the sensor 200 is also input to the second target motion specification prediction unit 20, and the movement specification of the second target output from the update processing unit 4 is established.
  • This embodiment differs from the configuration of the first embodiment in that the distribution 4 b is also input to the first target motion specification prediction unit 10.
  • the first target motion specification prediction unit 10 receives the detection signal 201 from the sensor 200 and, in addition to the motion specification probability distribution 4 a of the first target in the previous time frame from the update processing unit 4, the motion data of the second target The source probability distribution 4b is received, and the predicted value 10a of the motion specification probability distribution of the first target in the current time frame is sent to the update processing unit 4.
  • the second target motion specification prediction unit 20 receives the detection signal 201 from the sensor 200, and receives from the update processing unit 4 the motion specification probability distribution 4b of the second target in the previous time frame, and sends the update processing unit 4 to the current time.
  • the predicted value 20a of the movement specification probability distribution of the second target in the frame is sent.
  • the update processing unit 4 detects the detection signal 201 from the sensor 200, the predicted value 10a of the motion parameter probability distribution of the first target in the current time frame from the first target motion parameter predicting unit 10, and the second target motion data.
  • the predicted value 20a of the movement specification probability distribution of the second target in the current time frame from the original prediction unit 20 is received.
  • the update processing unit 4 calculates the motion specification probability distribution 4a of the first target in the current time frame, the motion specification probability distribution 4b of the second target in the current time frame, and the existence probability 4c of the first target in the current time frame.
  • the existence probability 4d of the second target in the current time frame is sent to the estimated value calculation unit 5.
  • the update processing unit 4 sends the first target motion specification probability distribution 4 a of the first target in the current time frame to the first target motion specification prediction unit 10 as the motion target probability of the first target in the previous time frame. Further, the update processing unit 4 sets the first target motion specification prediction unit 10 and the second target motion specification prediction unit 10 as the second target motion specification probability distribution 4b in the current time frame as the motion specification probability of the second target in the previous time frame. It sends to the target motion specification prediction unit 20. Furthermore, the update processing unit 4 determines the existence probability 4c of the first target in the current time frame and the existence probability 4d of the second target in the current time frame, the existence probability of the first target in the previous time frame, and the second target in the previous time frame. It sends to the existing probability predicting unit 3 as the existing probability of.
  • the hardware configuration of the target tracking device is realized by the configuration shown in FIG. That is, in the first target motion specification prediction unit 10 to the estimated value calculation unit 5 in FIG. 10, the program corresponding to each function is stored in the recording device 102, and the program corresponding to the processing is read by the arithmetic device 101. It is realized by being executed.
  • the configuration of each part is the same as that of the first embodiment, and thus the description thereof is omitted.
  • the motion specification of the estimation target is the position and velocity in the three-dimensional position space
  • the coordinate system is an orthogonal coordinate system
  • the motion of the target is assumed to be straight at a constant velocity
  • the detection signal 201 is the position of the detected signal.
  • the value be an intensity
  • the motion specification probability distribution be approximated by a linear combination of Gaussian distribution.
  • the first target motion specification prediction unit 10 uses the detection signal 201, the first target motion specification probability distribution at time k-1, and the second target motion specification probability distribution at time k-1.
  • the predicted value of the first target motion specification probability distribution at time k is calculated.
  • the Gaussian distribution constituting the predicted value of the first target motion specification probability distribution at time k is classified into three types.
  • the first is a Gaussian distribution representing the possibility that "the first target appears in the past from time k and does not combine with the second target at time k", and the first target motion specification probability at time k-1 Calculated from the distribution.
  • the second is a Gaussian distribution representing the possibility that "the first target appears at time k", and is calculated from the detection signal 201 of the past time frame from time k.
  • the third is a Gaussian distribution representing the possibility that "the first target appears in the past from time k and combines with the second target at time k", and the first target motion specification probability at time k-1 It is calculated from the distribution and the second target motion specification probability distribution at time k-1.
  • the Gaussian distribution representing the possibility that the first target "the first target appears in the past from time k and does not combine with the second target at time k" is suffixed with "S”
  • a Gaussian distribution that represents the possibility of the second "the first target appeared at time k" is denoted by a suffix "B”
  • the Gaussian distribution representing the possibility of the third "The first target appears in the past from time k and combines with the second target at time k" is given the suffix "M" It represents.
  • k ⁇ 1 Gaussian distributions representing the possibility that “the first target appears in the past from time k and does not combine with the second target at time k”.
  • the weighting factor, the Gaussian distribution mean value, and the Gaussian distribution covariance matrix are calculated from the following equations.
  • j 1, 2,..., J k ⁇ 1
  • k k and Q k are the same as the definition in step ST1 in the first embodiment.
  • w init , ⁇ k-1 and P init are the same as the definition in step ST1 in the first embodiment.
  • k-1 (1, jM) represents a Gaussian distributed covariance matrix after combining at time k-1 It is.
  • c 1 is a scalar value representing the magnitude of the influence of the first target's motion parameters before binding on the motion parameters after binding at time k-1
  • c 2 is a value at time k-1 It is a scalar value representing the magnitude of the influence exerted by the movement specification of the second target before combination on the movement specification after combination.
  • c 1 and c 2 are parameters, and set, for example, the assumed mass of the first target and the second target.
  • k-1 Gaussian distributions Is the predicted value of the first target motion specification probability distribution at time k (the symbol ⁇ represents a union). Each weighting coefficient in the predicted value of the first target motion specification probability distribution is standardized at the end of this step so as to satisfy the following equation.
  • the second target motion specification prediction unit 20 generates the second target motion specification probability distribution at time k from the detection signal 201 and the second target motion specification probability distribution at time k-1. Calculate the predicted value.
  • the Gaussian distribution that constitutes the predicted value of the second target motion specification probability distribution at time k is classified into two types. One is a Gaussian distribution representing the possibility that "the second target has appeared in the past from time k", and is calculated from the second target motion specification probability distribution at time k-1. The other is a Gaussian distribution representing the possibility that "the second target has appeared at time k", and is calculated from the detection signal 201 in the past time frame from time k.
  • the Gaussian distribution representing the possibility that "the second target has appeared in the past from time k" is suffixed with "S”
  • a Gaussian distribution that represents the possibility of "the second target appeared at time k" is It represents.
  • k-1 (2) Gaussian distributions representing the possibility that “the second target has appeared in the past from time k”, and each weighting coefficient, Gaussian distribution average value, Gaussian
  • the distribution covariance matrix is calculated from the following equation.
  • j 1, 2,..., J k ⁇ 1
  • k k and Q k are the same as the definition in step ST1 in the first embodiment.
  • w init , ⁇ k-1 and P init are the same as the definition in step ST1 in the first embodiment.
  • Gaussian distribution representing the possibility that "the first target has appeared in the past from time k" calculated by the above formulas (87) to (92), and "the first target has appeared in time k"
  • Gaussian distributions which is a combination of Gaussian distributions representing the possibility of Is the predicted value of the second target motion specification probability distribution at time k (the symbol ⁇ represents a union).
  • each weighting coefficient in the predicted value of the second target motion specification probability distribution is standardized at the end of this step so as to satisfy the following equation.
  • the predicted value of the exercise specification probability distribution of the first target at the current time is set to the exercise specification probability distribution of the first target at the previous time, It is configured to calculate also from the movement specification probability distribution of the second target at the current time.
  • the predicted values when the targets based on the law of inertia are combined, that is, when the first target and the second target are combined, the movement specifications such as the position and velocity of the first target immediately after combining are A predicted value of the motion parameter probability distribution of the first target according to the physical law that depends on the motion specifications of the first target and the second target is obtained.
  • the estimation accuracy of the motion specifications of the first target estimated by the update processing unit 4 and the estimated value calculation unit 5 based on the predicted values is, for example, the motion of each target as disclosed in the documents 1 and 4.
  • specification predicted values are estimated independently from movement target values of another target, it is an estimated value obtained from predicted values based on a model close to real physical laws, so true movement specification
  • the error with is small. That is, even if the observation time from coupling is short, or even if the number of times of detection of the first target after coupling is small due to the missing detection 202, a more accurate post-coupling than in the techniques disclosed in reference 1 and reference 4 It is possible to obtain the motion specifications of the first goal of the
  • the second target is a target which is different from the first target exercise specification probability distribution of the past time and the first target of the past time.
  • a first target motion specification prediction unit that predicts a motion specification probability distribution of the first target at the current time based on the motion specification probability distribution and target detection signal information that is observation information of the target;
  • a second target exercise specification prediction unit for predicting an exercise specification probability distribution of the second target at the current time based on the exercise specification probability distribution of the second target and target detection signal information, and a first target of the past time Based on the probability that the first target exists in the observation area and the second target of the past time exist in the observation area, the predicted value of the probability that the first target exists in the observation area at the current time and the current time
  • An existing probability predicting unit that calculates a predicted value of the probability that the second target is present in the observation area;
  • Motion specification probability distribution of the first target and the second target at the current time based on the output of the target motion specification prediction unit, the second target motion specification prediction unit,
  • the existence probability prediction unit monitors the first target at the current time according to the probability transition model between states regarding the presence or absence of the first target and the presence or absence of the second target. Since the predicted value of the probability existing in the area and the predicted value of the probability that the second target exists in the observation area at the current time are calculated, the motion specification of each target can be estimated with a small amount of calculation. it can.
  • the present invention allows free combination of each embodiment, or modification of any component of each embodiment, or omission of any component in each embodiment. .
  • the target tracking device estimates the number of targets and the movement specification of each target from the received signal of a sensor such as a radar or a camera for observing the target such as an aircraft or a flying object. It is suitable for estimating the motion specifications of each target when the targets are separated or combined.

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

Selon la présente invention, une première unité de prédiction de caractéristique de déplacement de cible (1) délivre en sortie des valeurs prédites (1a) pour une distribution de probabilité de caractéristique de déplacement pour une première cible. Une seconde unité de prédiction de caractéristique de déplacement de cible (2) délivre en sortie des valeurs prédites (2a) pour une distribution de probabilité de caractéristique de déplacement pour une seconde cible. Une unité de prédiction de probabilité de présence (3) délivre en sortie une valeur prédite (3a) pour une probabilité de présence pour la première cible et une valeur prédite (3b) pour une probabilité de présence pour la seconde cible. Une unité de traitement de mise à jour (4) met à jour ces valeurs prédites et délivre en sortie la distribution de probabilité de caractéristique de déplacement pour la première cible (4a), la distribution de probabilité de caractéristique de déplacement pour la seconde cible (4b), la probabilité de présence (4c) pour la première cible, et la probabilité de présence (4d) pour la seconde cible à une unité de calcul de valeur estimée (5). Sur la base de ces valeurs, l'unité de calcul de valeur estimée (5) calcule des valeurs de caractéristique de déplacement estimées (5a).
PCT/JP2017/025118 2017-07-10 2017-07-10 Dispositif de suivi de cible Ceased WO2019012575A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/JP2017/025118 WO2019012575A1 (fr) 2017-07-10 2017-07-10 Dispositif de suivi de cible
JP2019529328A JP6570800B2 (ja) 2017-07-10 2017-07-10 目標追尾装置

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2017/025118 WO2019012575A1 (fr) 2017-07-10 2017-07-10 Dispositif de suivi de cible

Publications (1)

Publication Number Publication Date
WO2019012575A1 true WO2019012575A1 (fr) 2019-01-17

Family

ID=65002171

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2017/025118 Ceased WO2019012575A1 (fr) 2017-07-10 2017-07-10 Dispositif de suivi de cible

Country Status (2)

Country Link
JP (1) JP6570800B2 (fr)
WO (1) WO2019012575A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2022175094A (ja) * 2021-05-12 2022-11-25 株式会社Soken 追跡装置
CN118200848A (zh) * 2024-05-16 2024-06-14 中国人民解放军海军工程大学 融合运动特征和探测信号的目标位置估计方法及电子设备

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH11271437A (ja) * 1998-03-20 1999-10-08 Mitsubishi Electric Corp 目標追尾装置及び目標追尾方法
JP2000180542A (ja) * 1998-12-10 2000-06-30 Mitsubishi Electric Corp 目標追尾装置
JP2004037262A (ja) * 2002-07-03 2004-02-05 Mitsubishi Electric Corp 目標追尾装置
JP2008261739A (ja) * 2007-04-12 2008-10-30 Mitsubishi Electric Corp 多目標追尾装置
JP2012127890A (ja) * 2010-12-17 2012-07-05 Mitsubishi Electric Corp 推定装置及びコンピュータプログラム及び推定方法
US20160161606A1 (en) * 2014-12-08 2016-06-09 Northrop Grumman Systems Corporation Variational track management

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH11271437A (ja) * 1998-03-20 1999-10-08 Mitsubishi Electric Corp 目標追尾装置及び目標追尾方法
JP2000180542A (ja) * 1998-12-10 2000-06-30 Mitsubishi Electric Corp 目標追尾装置
JP2004037262A (ja) * 2002-07-03 2004-02-05 Mitsubishi Electric Corp 目標追尾装置
JP2008261739A (ja) * 2007-04-12 2008-10-30 Mitsubishi Electric Corp 多目標追尾装置
JP2012127890A (ja) * 2010-12-17 2012-07-05 Mitsubishi Electric Corp 推定装置及びコンピュータプログラム及び推定方法
US20160161606A1 (en) * 2014-12-08 2016-06-09 Northrop Grumman Systems Corporation Variational track management

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2022175094A (ja) * 2021-05-12 2022-11-25 株式会社Soken 追跡装置
JP7593877B2 (ja) 2021-05-12 2024-12-03 株式会社Soken 追跡装置
CN118200848A (zh) * 2024-05-16 2024-06-14 中国人民解放军海军工程大学 融合运动特征和探测信号的目标位置估计方法及电子设备

Also Published As

Publication number Publication date
JPWO2019012575A1 (ja) 2019-11-07
JP6570800B2 (ja) 2019-09-04

Similar Documents

Publication Publication Date Title
Kirubarajan et al. Probabilistic data association techniques for target tracking in clutter
Gilholm et al. Spatial distribution model for tracking extended objects
Niedfeldt et al. Multiple target tracking using recursive RANSAC
US10935653B2 (en) Multi-target tracking method and tracking system applicable to clutter environment
CN114144697A (zh) 用于内插虚拟孔径雷达跟踪的系统和方法
US11486966B2 (en) Method and apparatus for tracking target from radar signal using artificial intelligence
EP3982160A1 (fr) Procédé et système de reconnaissance de fantômes à trajets multiples en intérieur
CN114708257B (zh) Sar运动舰船目标检测方法及装置
CN117197491B (zh) 基于高斯过程的概率假设密度滤波多机动目标跟踪方法
Kalandros et al. Multisensor covariance control strategies for reducing bias effects in interacting target scenarios
JP2020507767A (ja) 車両レーダシステム用の逆合成開口レーダ
US9921306B1 (en) Active optimal reduced state estimator
JP6570800B2 (ja) 目標追尾装置
Xu et al. FollowUpAR: Enabling follow-up effects in mobile AR applications
US20180372857A1 (en) Target tracking apparatus
CN113313739B (zh) 目标跟踪方法、装置和存储介质
US12380589B2 (en) Method and apparatus to evaluate radar images and radar device
CN116012412A (zh) 目标跟踪方法、装置、设备以及存储介质
Le et al. Human detection and tracking for autonomous human-following quadcopter
Petetin et al. Marginalized particle PHD filters for multiple object Bayesian filtering
JP6952914B2 (ja) 運動状態判定装置
JP7281472B2 (ja) 追尾対象識別装置、追尾対象識別方法、および、追尾対象識別プログラム
Javadi et al. Multi-target tracking pipeline for MIMO-FMCW radars based on modified GM-PHD
CN116758261A (zh) 基于生成对抗网络的宽带前视成像雷达目标识别方法
Lee et al. Enhanced Multiple Target Tracking Using a Generalized Multi-Target Smoothing Algorithm With Tracklet Association

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: 17917623

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2019529328

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17917623

Country of ref document: EP

Kind code of ref document: A1