Tracking device, tracking system, tracking method, and tracking method program

Abstract

To obtain a tracking device or the like capable of correctly tracking multiple tracked objects.SOLUTION: Based on a likelihood function by signals pertaining to multiple tracked objects, each of multiple tracking parts corresponding to the multiple tracked objects has an orientation change rate estimation part 509 for calculating an orientation change rate estimation value for estimating an orientation change rate in orientation of a corresponding tracked object, a weight generation part 510 for generating an orientation change rate weight based on an orientation change rate on the basis of an orientation change rate estimation value, a weight granting part 502 for generating a weight likelihood function of superposing an orientation change rate weight to the likelihood function, a Bayes estimation calculation part 503 for performing processing for generating a posterior probability distribution from the weight likelihood function, a Markov update part 507 for performing processing for generating a posterior probability distribution, a detection determination part 504 for performing processing for detecting a tracked object on the basis of a posterior probability distribution, an ID granting part 505 for granting an ID number to a tracked object, and an orientation estimation part 506 for estimating orientation of a tracked object on the basis of a posterior probability distribution.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] This technology relates to a tracking device, a tracking system, a tracking method, and a tracking method program that track a tracking target based on a signal, and in particular to tracking of multiple tracking targets. [Background technology]

[0002] Conventionally, there is known a tracking device that estimates a state quantity indicating the presence of a tracked target in a state quantity space based on a signal emitted by the object to be tracked in a noisy environment, and tracks the target based on the estimation. For example, the tracking device performs tracking by using a sound wave emitted by the tracked target as a signal and estimating the direction of arrival of the sound wave and the time change in frequency, etc. The state quantity indicating the presence of a tracked target corresponds to the direction of arrival of the sound wave and the frequency of the sound wave (for example, see Non-Patent Document 1). [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] REBethel and GJParas, “A PDF Multitarget Tracker”, IEEE Transactions on Aerospace and Electronic Systems, vol. 30, NO. 2, April 1994, p. 386-403 Summary of the Invention [Problem to be solved by the invention]

[0004] When there are multiple tracking targets, multiple trackers track each of them. Here, if the orientations of the multiple tracking targets are close to each other, each tracker cannot automatically determine which tracking target the estimated orientation obtained at a given time corresponds to. As a result, there is a possibility that multiple trackers tracking different tracking targets may estimate the orientation of the same tracking target and track it after the multiple tracking targets cross each other in terms of orientation.

[0005] Therefore, there has been a demand for a tracking device that can accurately track multiple tracking targets. [Means for solving the problem]

[0006] The tracking device according to the disclosure detects signals related to a plurality of tracking targets from observed value data. In multiple directions Calculate the likelihood and the direction are the parameter variables a weighting unit that generates a weighted likelihood function by superimposing the orientation change rate weight on the likelihood function; a Bayesian estimation unit that generates a posterior probability distribution from the weighted likelihood function; a Markov update unit that generates a priori probability distribution; a detection determination unit that determines the presence or absence of a tracking target based on the posterior probability distribution and detects the tracking target; an ID assignment unit that assigns an ID number to the tracking target detected by the detection determination unit; and an ID assignment unit that estimates the orientation of the tracking target based on the posterior probability distribution and assigns an estimated orientation. of and a direction estimation unit that calculates the direction.

[0007] In addition, the tracking system according to this disclosure comprises a receiver array having a plurality of receivers and a signal conversion device that converts signals received by the receivers into digital data to generate observation value data, and the tracking device described above.

[0008] The tracking method according to the present disclosure also includes a method for detecting a target object, the method being calculated from observed value data obtained by observing signals relating to a plurality of tracking targets. Multiple Pertaining to direction The orientation generated by the likelihood is used as a parameter variable. A tracking method for tracking each tracking target based on a likelihood function, comprising: a step of calculating a direction change rate estimated value by performing a direction change rate estimation process to estimate a direction change rate in the direction of the corresponding tracking target; a step of generating a direction change rate weight for weighting based on the direction change rate estimated value, a step of generating a weighted likelihood function by superimposing the direction change rate weight on the likelihood function; a step of generating a posterior probability distribution from the weighted likelihood function; a step of generating a priori probability distribution; a step of detecting the tracking target by determining the presence or absence of the tracking target based on the posterior probability distribution; of and a step of calculating the value.

[0009] The tracking method program according to the present disclosure also includes a tracking method for detecting a target object, the tracking method program being calculated from observation data obtained by observing signals related to a plurality of tracking targets. Multiple Pertaining to direction The orientation generated by the likelihood is used as a parameter variable. A tracking method program for tracking each tracking target based on a likelihood function, comprising: a step of calculating a direction change rate estimated value by performing direction change rate estimation processing to estimate a direction change rate in the direction of the corresponding tracking target; a step of generating a direction change rate weight for weighting based on the direction change rate estimated value, a step of generating a weighted likelihood function by superimposing the direction change rate weight on the likelihood function; a step of generating a posterior probability distribution from the weighted likelihood function; a step of generating a priori probability distribution; a step of determining the presence or absence of a tracking target based on the posterior probability distribution and detecting the tracking target; a step of assigning an ID number to the detected tracking target; and a step of estimating the direction of the tracking target based on the posterior probability distribution and calculating the estimated direction. ofThe calculation step is performed by a computer. [Effects of the Invention]

[0010] According to this disclosure, a weight generation unit generates an orientation change rate weight based on an orientation change rate estimated value calculated by an orientation change rate estimation unit. Then, a weighting unit superimposes the orientation change rate weight on a likelihood function, and a Bayesian estimation calculation unit calculates a posterior probability distribution based on the weighted likelihood function weighted based on the orientation change rate related to the orientation change, and the detection / determination unit and the orientation estimation unit detect and estimate the orientation of the tracking target. This makes it possible to calculate a probability distribution including orientation changes, and to accurately track multiple tracking targets. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a diagram showing the configuration of a tracking system centered around a tracking device 100 according to a first embodiment. [Figure 2] FIG. 2 is a diagram showing a configuration of a tracking unit 500 according to the first embodiment. [Figure 3] FIG. 3 is a diagram showing an example of a mask used in mask processing by a mask processing unit 501 according to the first embodiment. [Figure 4] FIG. 10 is a diagram illustrating weight processing in a weight generation unit 510 according to the first embodiment. [Figure 5] FIG. 2 is a diagram illustrating an example of a processing flow in the data processing device 110 according to the first embodiment. [Figure 6] FIG. 1 illustrates a situation where the orientations of two targets intersect. [Figure 7] FIG. 10 is a diagram showing an example of time transition of the azimuth of a target tracked by each tracking unit 500 when multiple targets intersect. [Figure 8] FIG. 10 is a diagram showing the mask likelihood function lm,it(θ) in each tracking unit 500 when erroneous tracking occurs. [Figure 9] FIG. 10 is a diagram illustrating the effect according to the first embodiment. [Figure 10] FIG. 10 is a diagram showing a configuration of a tracking unit 500 according to a second embodiment. [Figure 11] FIG. 11 is a diagram showing the configuration of a tracking unit 500 according to a third embodiment. [Figure 12] FIG. 11 is a diagram illustrating the processing of a peak likelihood extracting section 512 in the third embodiment. [Figure 13] FIG. 11 is a diagram illustrating the processing of an applied likelihood recalculation unit 513 in the third embodiment. [Figure 14] FIG. 11 is a diagram illustrating a regenerated likelihood function processed by a likelihood function regeneration unit 514 in the third embodiment. [Figure 15] 10A and 10B are diagrams illustrating the effects of the tracking device 100 according to the third embodiment. [Figure 16] FIG. 10 is a diagram showing the configuration of a tracking unit 500 according to a fourth embodiment. [Figure 17] 10A and 10B are diagrams illustrating the effects of the tracking device 100 according to the fourth embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, a tracking device according to an embodiment will be described with reference to the drawings. In the following drawings, components with the same reference numerals are identical or equivalent, and are common throughout the following embodiments. Furthermore, the size relationships between components in the drawings may differ from those in reality. Furthermore, the configurations of components shown throughout the specification are merely examples and are not limited to the configurations described in the specification. Not all of the devices described in the specification may be included. In particular, the combinations of components are not limited to the combinations in each embodiment, and components described in other embodiments may be applied to other embodiments. Furthermore, when multiple similar devices are distinguished by subscripts, the reference numerals, subscripts, etc. may be omitted if there is no need to distinguish or identify them.

[0013] Embodiment 1 FIG. 1 is a diagram showing the configuration of a tracking system centered on a tracking device 100 according to a first embodiment. In explaining the tracking system according to the first embodiment, a sound wave signal generated by sound waves will be used as an example of a signal used to track a target to be tracked. Hereinafter, the tracked target that serves as a sound source will be referred to as a target. The sound wave signal includes sound waves that serve as acoustic signals obtained from the target as well as sound waves that serve as noise. Here, the tracking system according to the first embodiment will be explained assuming a case in which two target A and target B that serve as sound sources are tracked.

[0014] The tracking system according to the first embodiment includes a tracking device 100, a microphone array 200, and a communication line 300. As shown in FIG. 1, the microphone array 200 is a receiver array in which a plurality of microphones 210-1 to 210-n, which convert received sound signals into electrical signals, are arranged in an arbitrary shape, such as a line. The microphone array 200 also includes a signal conversion device 220 and a signal transmission device 230. The signal conversion device 220 periodically performs sampling and encoding processes on the sound signals received by each microphone 210, converting them into observation value data, which is digitalized data to be processed, and performs data generation processing. The signal transmission device 230 transmits a signal including the observation value data.

[0015] The communication line 300 is a line that connects the microphone array 200 and the tracking device 100 for communication. Here, the communication line 300 is assumed to be a dedicated line for communication, but is not limited to this. For example, a signal may be sent from the microphone array 200 to the tracking device 100 via a commonly used electric communication line. The communication line 300 may be wired or wireless.

[0016] Tracking device 100 is a device that performs processing related to target tracking based on observation data. Tracking device 100 in embodiment 1 includes data processing device 110, storage device 120, signal input device 130, and signal output device 140. Based on a signal sent from microphone array 200 via communication line 300, signal input device 130 converts data in the signal into observation data that can be processed by data processing device 110. Furthermore, signal output device 140 generates a signal including data on the data processing and detection determination results, and transmits the signal to an external device (not shown). Here, in embodiment 1, tracking device 100 includes signal input device 130 and signal output device 140, but devices other than tracking device 100 may include these.

[0017] Furthermore, the storage device 120 included in the tracking device 100 in the first embodiment temporarily or long-term stores data used for processing by the data processing device 110. Here, as will be described later, the data processing device 110 has a plurality of tracking units 500. Here, the storage device 120 collectively stores data, parameters, etc. for the plurality of tracking units 500. However, this is not limited to this. An independent storage device 120 may be provided for each tracking unit 500. Here, values related to the data, parameters, etc. required for each tracking unit 500 to perform its respective processing may differ.

[0018] The data processing device 110 included in the tracking device 100 of the first embodiment calculates estimated azimuths θ of a plurality of targets based on the observation value data related to the sound wave signals. * and generates and processes data on the ID numbers assigned to each target. The data processing device 110 has a likelihood generation unit 400 and multiple tracking units 500. In FIG. 1, the data processing device 110 has two tracking units 500-1 and 500-2. Each tracking unit 500 basically has the same configuration and performs the same processing. Each unit within the data processing device 110 performs processing such as calculation and judgment, and generates various data for processing by the data processing device 110.

[0019] Here, the data processing device 110 of the tracking device 100 is typically configured with a device that performs control and arithmetic processing, such as a computer centered around a CPU (Central Processing Unit).The data processing device 110 typically executes a pre-programmed signal processing procedure performed by each unit to realize the processing of each unit.Here, for example, the storage device 120 stores the program data.However, this is not limited to this, and each unit may be configured with a separate dedicated device (hardware).

[0020] The storage device 120 also has a volatile storage device (not shown) such as a random access memory (RAM) that can temporarily store data, and a non-volatile auxiliary storage device (not shown) such as a hard disk drive. The non-volatile auxiliary storage device may be a solid state drive, a flash memory that can store data for a long period of time, or the like. Here, the tracking device 100 may be configured such that the hardware that executes the processing of each unit of the data processing device 110 and the storage device 120 is located in different positions, and the units communicate with each other to perform the processing operations of the tracking device 100. The tracking unit 500 may also be an independent device that functions as a tracker.

[0021] The likelihood generating unit 400 generates likelihoods for a plurality of directions based on the observation value data related to the sound wave signals received by the microphone array 200, and calculates a likelihood function l with the direction θ as a parameter variable. t (θ) is calculated. Here, t represents time (t=1, 2, ...), and the period from time t to time t+1 is the processing cycle of the data processing device 110. The tracking device 100 of the first embodiment will be described below as processing observed value data at a certain time t. The likelihood generating unit 400 forms directivities in a plurality of directions simultaneously by, for example, beamforming. The likelihood generating unit 400 then arranges and processes the power values of the sound in each direction, and calculates the obtained data using the likelihood function l t Here, the likelihood generator 400 calculates the likelihood function l using beamforming. t(θ) data is generated, but it is not limited to this. For example, by assuming a distribution of the target direction that is the sound source, the instantaneous estimated direction θ of the target can be calculated. * The likelihood function l for the neighborhood of t (θ) and calculate the likelihood function l t Other techniques for generating (θ) are available.

[0022] 2 is a diagram showing the configuration of the tracker 500 according to the first embodiment. Each tracker 500 calculates the likelihood function l generated by the likelihood generator 400. t (θ) and the estimated direction θ of the corresponding target. * and ID number as data. Each tracking unit 500 in the first embodiment includes a mask processing unit 501, a weighting unit 502, a Bayesian estimation calculation unit 503, a Markov update unit 507, a mask generation unit 508, a detection determination unit 504, an ID assignment unit 505, an orientation change rate estimation unit 509, a weight generation unit 510, and an orientation estimation unit 506.

[0023] 3 is a diagram showing an example of a mask used for masking by the mask processing unit 501 in the first embodiment. The mask processing unit 501 uses the likelihood function l generated by the likelihood generating unit 400. t A masking process is performed on (θ) to superimpose a mask as shown in FIG. 3(a) generated by a mask generating unit 508 (to be described later) to obtain a mask likelihood function l m,i t (θ) is generated, where i is a number that identifies the tracking unit 500. The masking process is performed by using a likelihood function l as shown in FIG. tAs shown in FIG. 3(c), this is a process of suppressing likelihoods other than those in a specific range (width θmask) of azimuth (θ). Here, the mask processing unit 501 performs processing using a mask that is stored in a mask storage unit 522 (described later) and that is generated by a mask generation unit 508 (described later) at time t-1, which is one cycle before time t. If the initial value of the mask used by the mask processing unit 501 during processing at time t=1 is not set in advance, the same value, for example, 1, is superimposed on all azimuths so that the relationship between azimuth and likelihood does not change due to the mask. On the other hand, if the azimuth, etc. are set in advance, the initial value of the mask is set to a value based on the mask set in advance. The mask likelihood function l m,i t (θ) is used as data in the processing performed by weighting unit 502.

[0024] The weighting unit 502 weights the masked likelihood function l generated by the mask processing unit 501 through the masking process. m,i t (θ) is the orientation change rate weight w generated by the weight generation unit 510 described later. t (θ') is weighted to obtain the weighted likelihood function L w,i t (θ,θ') is generated. The weighted likelihood function L w,i t (θ, θ') is used as data in the processing performed by the Bayesian estimation calculation unit 503. Here, θ' represents the heading change rate. The weighted likelihood function L w,i t (θ, θ') is a vector with the direction θ and the direction change rate θ' as parameters (variables).

[0025] The Bayesian estimation calculation unit 503 calculates the weight likelihood function L related to the processing of the weighting unit 502. w,i t (θ, θ′) and the prior probability distribution P prior,i t|t-1 Calculation is performed based on (θ,θ'), and the posterior probability distribution P posterior,i t(θ, θ'). The Bayesian estimation calculation unit 503 calculates the posterior probability distribution P posterior,i t The posterior probability distribution P of the direction obtained by adding (θ, θ') to the direction of the direction change rate for each direction θ posterior,i t (θ) is calculated by the Bayesian estimation calculation unit 503. posterior,i t (θ, θ′) is used as data in the process performed by the Markov update unit 507. In addition, the orientation posterior probability distribution P posterior,i t (θ) is used as data in the processing performed by the direction estimation unit 506 and the detection determination unit 504.

[0026] The prior probability distribution P used by the Bayesian estimation calculation unit 503 during calculation processing prior,i t|t-1 (θ, θ′) is the prior probability distribution P at one period before time t (time t−1) stored in the prior probability buffer unit 521 described later. prior,i t|t-1 (θ, θ′). Here, the prior probability distribution P prior,i 1|0 Generally, when the initial value of (θ) is not set in advance, it is set to a value that has an equal probability of 1 / Ng for all the target orientations θ and orientation change rates θ', where Ng is the number of intervals in the state space in the prior probability distribution. On the other hand, when it is set in advance, the prior probability distribution P prior,i t|t-1 The initial values of (θ, θ') are set to values that result in an arbitrary distribution based on a pre-setting.

[0027] The detection determination unit 504 uses the posterior probability distribution P posterior,i t Based on (θ, θ'), a determination process is performed to determine whether or not there is a target that is a sound source.

[0028] The ID assigning unit 505 performs a process of assigning an ID number to the target determined by the detection and determination unit 504. Here, while the detection and determination unit 504 continuously determines that a target is present, the ID assigning unit 505 continues to assign the same ID number. Furthermore, when the detection and determination unit 504 that had determined that no target was present determines that a target is present, the ID assigning unit 505 assigns a new ID number. The ID number may be any identifiable number such as numbers, letters, or symbols.

[0029] The Markov update unit 507 updates the posterior probability distribution P posterior,i t Based on (θ,θ'), the prior probability distribution P prior,i t+1|t (θ, θ') is calculated. The Markov update unit 507 calculates the prior probability distribution P prior,i t+1|t (θ, θ') is used in the processing of the mask generation unit 508 and the orientation change rate estimation unit 509. In addition, the prior probability distribution P prior,i t+1|t (θ, θ′) is stored in the prior probability buffer unit 521, which will be described later. The prior probability distribution P prior,i t+1|t (θ, θ′) is the prior probability distribution P prior,i t|t-1 It is used as (θ,θ').

[0030] The mask generation unit 508 calculates the expected value of the orientation θ using the weighted average, and calculates the orientation θ with the maximum prior probability. t+1|t Then, the mask generating unit 508 estimates the target's direction θ at time t+1. t+1|t Based on this, a mask is generated in which a certain value (for example, 1) is set in the width θmask based on the orientation at time t+1, and a value lower than the certain value (for example, 0) is set in the other orientations. The mask data generated by the mask generation unit 508 is stored in a mask storage unit 522, which will be described later, and is used in the mask processing performed by the mask processing unit 501 in the next cycle.

[0031] The azimuth change rate estimation unit 509 estimates the azimuth change rate θ' for the target. * The azimuth change rate estimation process is performed to calculate the azimuth change rate.

[0032] The weight generation unit 510 calculates the heading change rate estimated value θ′ processed by the heading change rate estimation unit 509. * and the set adjustment parameters σ stored in the adjustment parameter storage unit 524. 2 θset Based on this, the heading change rate weight w used in the processing of the next period (time t+1) is t (θ') is calculated. t The data of (θ′) is stored in the weight storage unit 523 and is used in the weighting process performed by the weighting unit 502 in the next cycle.

[0033] The direction estimation unit 506 then calculates the posterior probability distribution P posterior,i t Based on (θ, θ'), the estimated direction θ is the direction of the sound source (target) from which the acoustic signal is emitted. * Here, the direction estimation process performed by the direction estimation unit 506 is a process that is performed using techniques such as MAP estimation and weighted averaging.

[0034] Furthermore, data required for processing in the tracker 500 is stored in the storage device 120. Here, it is assumed that corresponding data is stored in particular in the a priori probability buffer unit 521, the mask storage unit 522, the weight storage unit 523, and the adjustment parameter storage unit 524. Here, the a priori probability buffer unit 521, the mask storage unit 522, the weight storage unit 523, and the adjustment parameter storage unit 524 are included in the storage device 120, but for convenience of explanation, they will be described below as part of the tracker 500.

[0035] The prior probability buffer unit 521 stores the prior probability distribution P prior,i t+1|t(θ) for one period. As described above, the prior probability distribution P prior,i t+1|t (θ, θ′) is the prior probability distribution P prior,i t|t-1 It is used as (θ,θ').

[0036] The mask storage unit 522 stores one period's worth of masks generated by the mask generation unit 508. The masks stored in the mask storage unit 522 are used for mask processing by the mask processing unit 501 at time t+1.

[0037] The weight storage unit 523 stores the orientation change rate weight w generated by the weight generation unit 510. t (θ') for one period. The weight storage unit 523 stores the weight w t (θ′) is used in the processing of weighting unit 502 at time t+1.

[0038] The adjustment parameter storage unit 524 stores preset adjustment parameters σ 2 θset The setting adjustment parameter σ is stored as data. 2 θset is the heading change rate weight w obtained by the processing of the weight generation unit 510 in the previous cycle. t This is a parameter that adjusts the spread of the direction of the azimuth change rate θ' of (θ'). Here, the setting adjustment parameter σ 2 θset If is set to a small value, it becomes easier to distinguish the heading change rate θ' in the direction where two targets intersect. However, the estimation accuracy of the heading change rate θ' of each target decreases. Therefore, the setting adjustment parameter σ 2 θset The parameter value is set in advance, taking into consideration the trade-off between the heading change rate θ' and the estimation accuracy of the heading change rate θ'. For example, if emphasis is placed on the ability to distinguish between the heading change rates θ' of two targets, the prior probability distribution P prior,i t|t-1 (θ,θ') and the posterior probability distribution P posterior,i tThe step size of the direction change rate θ' in (θ,θ') is set as the adjustment parameter σ 2 θset Set.

[0039] Next, the processing of the data processing device 110 included in the tracking device 100 in the first embodiment will be described. Here, the processing of the tracking unit 500 at time t will be mainly described. Here, in order to vary the targets to be tracked, the initial values of the mask used by the mask processing unit 501 are different, but the same processing is performed in each tracking unit 500. First, the processing in the orientation change rate estimation unit 509 and the weight generation unit 510 at time t-1 used at time t will be described.

[0040] The heading change rate estimation unit 509 uses the prior probability distribution P prior,i t|t-1 Add (θ, θ') in the azimuth direction to obtain the azimuth change rate prior probability distribution P prior,i t|t-1 Then, the heading change rate estimation unit 509 calculates the heading change rate prior probability distribution P prior,i t|t-1 For (θ'), the estimated value of the heading change rate θ' at time t is calculated by calculating the expected value of the heading change rate θ' using the weighted average or by estimating the heading change rate θ' with the maximum prior probability. * Here, the heading change rate estimation unit 509 calculates the heading change rate prior probability distribution P prior,i t|t-1 (θ') and calculate the estimated heading change rate θ' * However, the prior probability distribution P prior,i t|t-1 Regarding (θ,θ'), in the space of the direction θ and the direction change rate θ', the direction change rate θ' is estimated by calculating the expected value of the direction change rate θ' using a weighted average, or by estimating the direction change rate θ' with the maximum prior probability, and the direction change rate estimated value θ' is obtained. * may be calculated.

[0041] 4 is a diagram illustrating weight processing by the weight generating unit 510 according to embodiment 1. The weight generating unit 510 uses the orientation change rate θ′ processed by the orientation change rate estimating unit 509 and the set adjustment parameter σ 2 θset Based on this, the heading change rate weight w t The weight generation unit 510 performs a weighting process to calculate the heading change rate weight w (θ′) based on the equation (1), as shown in FIG. t Generate (θ').

[0042]

number

[0043] 5 is a diagram illustrating an example of the flow of processing in the data processing device 110 according to the first embodiment. t The processing of each tracking unit 500 based on (θ) will be described. The likelihood generating unit 400 generates likelihoods in a plurality of directions based on the observation value data related to the acoustic signals received by the microphone array 200, as described above, and calculates likelihood function l t (θ) is calculated and generated (step S0).

[0044] The mask processing unit 501 calculates the likelihood function l generated by the likelihood generating unit 400. t (θ), based on the mask stored in the mask storage unit 522, the mask likelihood function l m,i t (θ) is generated (step S1).

[0045] The weighting unit 502 weights the masked likelihood function l generated by the mask processing unit 501 through the masking process. m,i t (θ) and the orientation change rate weight w generated by the weight generation unit 510 t (θ') and calculate the weighted likelihood function L w,i t (θ, θ') is generated (step S2).

[0046]

number

[0047] Here, the heading change rate weight w t (θ') is a vector related to the orientation change rate θ'. Also, the mask likelihood function l m,i t (θ) is a vector related to the direction θ. In equation (2), the result of multiplying the combination of the elements of all vectors related to the direction change rate θ' and the direction θ is the weighted likelihood function L w,i t (θ,θ').

[0048] The Bayesian estimation calculation unit 503 calculates the weight likelihood function L related to the processing of the weighting unit 502. w,i t (θ, θ′) and the prior probability distribution P prior,i t|t-1 Using (θ,θ'), the posterior probability distribution P posterior,i t (θ, θ′). Furthermore, the Bayesian estimation calculation unit 503 calculates the posterior probability distribution P posterior,i t The posterior probability distribution P of the direction obtained by adding (θ, θ') to the direction of the direction change rate for each direction θ posterior,i t (θ) is calculated (step S3).

[0049]

number

[0050]

number

[0051] The detection determination unit 504 calculates the posterior probability distribution P posterior,i tBased on (θ), a determination process is performed to determine whether or not there is a target that is a sound source (step S4). Then, the ID assigning unit 505 performs a process of assigning an ID number to the target determined by the detection determining unit 504 (step S5).

[0052] The direction estimation unit 506 calculates the direction posterior probability distribution P posterior,i t (θ) is used to estimate the direction of the sound source (target) from which the acoustic signal is emitted. * is generated (step S6).

[0053] Here, the detection determination unit 504 and the direction estimation unit 506 are configured to use the Bayesian estimation calculation unit 503 to calculate the posterior probability distribution P posterior,i t The posterior probability distribution P of the orientation calculated from (θ,θ') posterior,i The detection determination unit 504 and the direction estimation unit 506 perform the processing based on the posterior probability distribution P posterior,i t With respect to (θ, θ′), processing may be performed by detecting a peak in the space of the orientation θ and the orientation change rate θ′.

[0054] Next, the processing of the Markov update unit 507 will be described. The Markov update unit 507 updates the posterior probability distribution P posterior,i t Based on (θ,θ'), the prior probability distribution P prior,i t+1|t (θ, θ') is calculated and generated (step S7). prior,i t+1|t (θ, θ′) is stored in the prior probability buffer unit 521.

[0055] As described above, the heading change rate estimation unit 509 calculates the prior probability distribution P prior,i t+1|t Adding (θ, θ') in the azimuth direction, the azimuth change rate prior probability distribution P prior,it+1|t (θ') is calculated (step S8).

[0056] Then, the weight generation unit 510 calculates the azimuth change rate θ′ processed by the azimuth change rate estimation unit 509 and the set adjustment parameter σ 2 θset Based on this, weighting is performed, and the orientation change rate weight w t (θ') is generated (step S9).

[0057] 6 is a diagram showing a situation in which the directions of two targets intersect. As shown in Fig. 6, when viewed from the position of microphone array 200, the two targets that are sound sources move in opposite directions.

[0058] 7A and 7B are diagrams showing an example of time transition of the direction of a target tracked by each tracking unit 500 when multiple targets intersect. Fig. 7A shows the estimated direction θ of the target by each tracking unit 500 when the target is being correctly tracked. * 7(b) shows the estimated azimuth θ of the target by each tracking unit 500 when erroneous tracking occurs. * In Figure 7, a shows a situation where targets A and B, which are separated by different directions, gradually approach each other. b shows a situation where the two targets pass each other in terms of direction and their directions overlap. c shows a situation where the two targets gradually move away from each other after passing each other.

[0059] FIG. 8 shows the mask likelihood function l in each tracking unit 500 when a tracking error occurs. m,i t 8(a) is a diagram showing the mask likelihood function l(θ) in the tracker 500-1 and the tracker 500-2 in the situation a in FIG. m,i t 8(b) shows the mask likelihood function l in the tracker 500-1 and the tracker 500-2 in the situation b in FIG. m,i t8(c) shows the mask likelihood function l in the tracker 500-1 and the tracker 500-2 in the situation c in FIG. m,i t As shown in Figure 8(a), when the target is far away, the mask likelihood function l m,i t In (θ), the direction of the peak likelihood for each target is different. On the other hand, the mask likelihood function l in Fig. 8(b) m,i t Regarding (θ), the likelihood of two targets is included, and the likelihood peak cannot be reduced to one.

[0060] The Bayesian estimation calculation unit 503 calculates the mask likelihood function l in FIG. m,i t As a result of performing processing based on (θ), the peak in the posterior probability distribution in tracker 500-2 will be at the position of target A. The prior probability distribution calculated by Markov updater 507, which performs processing using the posterior probability distribution, will also have a peak at target A. As a result, as shown in FIG. 7(b), after the targets cross, tracker 500-2 will start tracking target A, the same as tracker 500-1, instead of target B.

[0061] 9 is a diagram illustrating the effect according to Embodiment 1. According to the tracking device 100 of Embodiment 1, the weight generation unit 510 calculates the azimuth change rate estimated value θ′ obtained by the azimuth change rate estimation unit 509. * From the heading change rate weight w t Then, the weighting unit 502 calculates the azimuth change rate weight w t The weighted likelihood function L is weighted based on (θ'). w,i t Generate (θ,θ').

[0062] As shown in Fig. 9(a), in the situation a shown in Fig. 7, the direction of the target A changes in the positive direction. Therefore, the weighted likelihood function L w,i tFor (θ, θ'), the likelihood of a positive azimuth change rate θ' is high. On the other hand, the azimuth of target B changes in the negative direction. Therefore, the weighted likelihood function L w,i t (θ, θ') has a high likelihood of a negative heading rate θ'.

[0063] 9(b), in the situation b shown in FIG. 7, both the tracking unit 500-1 and the tracking unit 500-2 use a weighted likelihood function L w,i t (θ, θ') is generated for each azimuth change rate θ'. As a result, in each tracking unit 500, the posterior probability distribution P posterior,i t (θ, θ') also reflects the respective azimuth change rates θ' for target A and target B.

[0064] 7, the azimuths θ of the targets A and B become separated over time. Therefore, as shown in FIG. 9(c), the weighted likelihood function L w,i t (θ, θ') reflects the azimuth θ and azimuth change rate θ' of the target being tracked. Therefore, tracking device 100 of the tracking system in embodiment 1 can accurately track multiple targets even when multiple targets intersect.

[0065] Embodiment 2 Fig. 10 is a diagram showing the configuration of tracking unit 500 according to embodiment 2. In Fig. 10, mask processing unit 501, weighting unit 502, Bayesian estimation calculation unit 503, Markov update unit 507, mask generation unit 508, detection determination unit 504, ID assignment unit 505, and orientation change rate estimation unit 509 perform the same processing as that described in embodiment 1.

[0066] Here, in the tracking device 100 of the second embodiment, the estimated direction θ obtained by the direction estimation unit 506 through the estimation process is *is not only output as a signal but also stored as data in the estimated azimuth buffer unit 526, which will be described later. In addition, the weight generation unit 510 calculates the azimuth change rate weight w to be used in the processing of the next period (time t+1) based on the error standard deviation σθ obtained by processing in the error standard deviation calculation unit 511, which will be described later, and the adjustment parameter k stored in the adjustment parameter storage unit 524. t In the second embodiment, the adjustment parameter k stored in the adjustment parameter storage unit 524 is calculated by multiplying the error standard deviation σθ related to the heading change rate θ′ by the heading change rate weight w t This parameter adjusts the degree to which the spread of (θ') is affected.

[0067] 10, in the second embodiment, each tracking unit 500 of the tracking device 100 has an error standard deviation calculation unit 511. The tracking device 100 of the second embodiment also has an estimated orientation change rate buffer unit 525 and an estimated orientation buffer unit 526 in the storage device 120.

[0068] The estimated heading change rate buffer unit 525 stores the heading change rate θ′ obtained by the heading change rate estimation unit 509 as data for one period. θ+1 T from time t θ The rate of change of azimuth θ' for one period is stored as data.

[0069] The estimated direction buffer unit 526 stores the estimated direction θ obtained by the direction estimation process of the direction estimation unit 506. * Here, the estimated direction buffer unit 526 stores the data at time tT θ T from time t θ+1 Estimated direction θ for the period * is stored as data.

[0070] The error standard deviation calculation unit 511 calculates the estimated azimuth θ stored in the estimated azimuth buffer unit 526. * and the heading change rate θ' stored in the estimated heading change rate buffer unit 525, and performs calculation processing based on equations (5) and (6), and calculates the error standard deviation σθ' for the heading change rate θ'.* Here, ΔT in equation (6) is calculated. proc is the estimated direction θ * The error standard deviation calculation unit 511 calculates the azimuth change rate θ′, which is the magnitude of the azimuth change per unit time, using equation (6).

[0071]

number

[0072]

number

[0073] The weight generation unit 510 performs calculations based on the equation (7) to generate the heading change rate weight w t (θ') is generated. The heading change rate weight w t The data of (θ') is stored in the weight storage unit 523, and in the next period, the heading change rate weight w t (θ′) is used in the weighting process performed by the weighting unit 502.

[0074]

number

[0075] For example, in the tracking device 100 of the first embodiment, the weight generation unit 510 uses the set adjustment parameters σ 2 θset The weight of the change in the direction ... t Determine the spread of (θ') where the setting adjustment parameter σ 2 θset The weight of the heading change rate w t If the spread of (θ') is set to be small, the azimuth change rate estimated value θ' *However, if the target's heading change rate θ' differs significantly from the actual value, the heading change rate θ' in subsequent calculations may be incorrect. On the other hand, the heading change rate weight w t If the spread of (θ') is set to be large, it becomes difficult to distinguish between multiple targets using the azimuth change rate θ', and there is a possibility that accurate tracking will not be possible when multiple targets intersect.

[0076] On the other hand, in the tracking device 100 of the second embodiment, the azimuth change rate weight w for the azimuth change rate θ′ is calculated in accordance with the product of the difference between the estimated value of the azimuth change rate θ′ and the actual azimuth change rate θ′ of the target that is the sound source and the adjustment parameter k. t (θ') can be automatically adjusted. Specifically, the greater the difference between the estimated heading rate θ' and the target's actual heading rate θ', the greater the heading rate weight w t (θ') becomes wider, and the smaller the difference between the estimated value of the heading rate θ' and the target's actual heading rate θ', the greater the heading rate weight w t (θ') can be automatically adjusted to narrow the range, resulting in highly accurate estimation of the heading rate θ' and accurate tracking of multiple targets.

[0077] Embodiment 3 Fig. 11 is a diagram showing the configuration of tracking unit 500 according to embodiment 3. In Fig. 11, the components denoted by the same reference numerals as in Fig. 10 perform the same processing as described in embodiment 1 and embodiment 2.

[0078] 11 , the tracking unit 500 of the tracking device 100 according to the third embodiment includes a peak likelihood extraction unit 512, an applied likelihood recalculation unit 513, and a likelihood function regeneration unit 514. The tracking unit 500 according to the third embodiment also includes a maximum likelihood buffer unit 527 and an orientation likelihood width parameter storage unit 528 in the storage device 120.

[0079] The maximum likelihood buffer unit 527 stores the maximum likelihood l extracted by the peak likelihood extracting unit 512 (to be described later). peak , the set time (time tTlmax T from time t lmax The azimuth likelihood width parameter storage unit 528 stores the azimuth likelihood width parameter as data. The azimuth likelihood width parameter is a parameter that adjusts the spread of the likelihood in the azimuth direction used by the Bayesian estimation calculation unit 503.

[0080] The peak likelihood extraction unit 512 extracts the masked likelihood function l generated by the mask processing unit 501. m,i t Among (θ), the maximum likelihood l peak The direction θ where the peak likelihood was extracted is called the maximum direction θ peak The maximum likelihood l extracted by the peak likelihood extraction unit 512 is peak is used as data in the process performed by the applied likelihood recalculation unit 513. In addition, the maximum likelihood l extracted by the peak likelihood extraction unit 512 peak is stored as data in maximum likelihood buffer unit 527.

[0081] The applied likelihood recalculator 513 calculates the maximum likelihood l for the set amount stored in the maximum likelihood buffer 527. peak and the maximum likelihood l obtained by the processing of the peak likelihood extraction unit 512. peak and maximum orientation θ peak The Bayesian estimation calculation unit 503 generates an application likelihood and a direction application corresponding to the application likelihood.

[0082] The likelihood function regeneration unit 514 recalculates the likelihood function based on the applied likelihood obtained by processing by the applied likelihood recalculation unit 513, the applied orientation, and the orientation likelihood width parameter stored in the orientation likelihood width parameter memory unit 528, to generate a regenerated likelihood function.

[0083] Next, a further description will be given of the processing of the tracker 500 in the tracking device 100 according to the third embodiment. Here, the processing of the peak likelihood extractor 512, the applied likelihood recalculator 513, and the likelihood function regenerator 514 will be particularly described.

[0084] The peak likelihood extractor 512 extracts the maximum likelihood l based on equation (8). peak and maximum orientation θ peak Here, the function findpeaks is a function that finds the maximum peak (maximum value). peak,i t ,θ peak,i t ] is the maximum likelihood l peak and maximum likelihood l peak The maximum azimuth θ corresponding to peak are stored in descending order. Then, the peak likelihood extractor 512 calculates [l peak,i t ,θ peak,i t ], the maximum likelihood l peak is stored as data in maximum likelihood buffer unit 527.

[0085]

number

[0086] 12A and 12B are diagrams illustrating the processing of the peak likelihood extractor 512 in the third embodiment. Fig. 12A shows the masked likelihood function l in the tracker 500 that tracks each target when multiple targets (two targets in this example) are in the state shown in Fig. 7A. m,i t (θ) and maximum likelihood l peak As shown in FIG. 12(a), if the directions of the two targets are different, the peak likelihood extractor 512 of the tracker 500-1 extracts the maximum likelihood l that is the peak related to the acoustic signal from the target A. peak,1 t Furthermore, the peak likelihood extraction unit 512 of the tracking unit 500-2 extracts the maximum likelihood l that is the peak related to the acoustic signal from the target B. peak,2 t (B) is extracted. Then, the maximum likelihood l obtained by the extraction of each tracking unit 500 is peak are stored as data in maximum likelihood buffer unit 527.

[0087] On the other hand, FIG. 12(b) shows the mask likelihood function l in the tracking unit 500 that tracks each target when multiple targets are in the state shown in FIG. 7(b). m,i t (θ) and maximum likelihood l peak As shown in FIG. 12(b), if the directions of two targets are close to each other, the tracking unit 500-1 calculates the maximum likelihood l peak,1 t (A) and maximum likelihood l peak,1 t (B) is extracted. The tracking unit 500-2 also extracts the maximum likelihood l peak,2 t (A) and maximum likelihood l peak,2 t Extract (B).

[0088] Next, the applied likelihood recalculator 513 calculates the maximum likelihood l for the set amount stored in the maximum likelihood buffer 527. peak and the maximum likelihood l obtained by the processing of the peak likelihood extraction unit 512. peak and maximum orientation θ peak First, the applied likelihood recalculator 513 calculates T lmax The mean value of the maximum likelihood of l,i t Calculate.

[0089]

number

[0090] The applied likelihood recalculator 513 further calculates the maximum likelihood l peak and maximum orientation θ peak and the average value μ l,i t Using this, the application likelihood l re_aplly,i t and the applied direction θ based on Eq. (11) re_aplly,i t Here, i in equations (10) and (11) is calculated. peak.apply,i is expressed by equation (12).

[0091]

number

[0092]

number

[0093]

number

[0094] 13 is a diagram illustrating the processing of applied likelihood recalculation section 513 in embodiment 3. As shown in the upper part of FIG. 13, for example, applied likelihood recalculation section 513 of tracker 500-1 calculates maximum likelihood l in time series for a set time stored in maximum likelihood buffer section 527. peak Based on the maximum likelihood l peak,1 t (A) and maximum likelihood l peak,1 t (B) Among them, the maximum likelihood l peak,1 t (A) and maximum likelihood l peak,1 t The maximum orientation θ corresponding to (A) peak can be generated.

[0095] On the other hand, as shown in the lower part of FIG. 13, the applied likelihood recalculator 513 of the tracker 500-2 calculates the maximum likelihood l based on the maximum likelihood for the set time stored in the maximum likelihood buffer 527. peak,2 t (B) and maximum likelihood l peak,2 t (B) The maximum orientation θ peak can be generated.

[0096] In this way, the applied likelihood recalculation unit 513 calculates the masked likelihood function l m,i t (θ) is not the magnitude of the maximum likelihood, but the maximum likelihood l that is closer to the past maximum likelihood stored in the maximum likelihood buffer unit 527. peak and maximum orientation θpeak As a result, in FIG. 13(b), the adaptive likelihood recalculator 513 of the tracker 500-1 generates [l peak,1 t ,θ peak,1 t ], and the applied likelihood recalculation unit 513 of the tracker 500-2 recalculates [l peak,2 t ,θ peak,2 t ] is recalculated.

[0097] 14A and 14B are diagrams illustrating the regenerated likelihood function processed by the likelihood function regenerator 514 in the third embodiment. FIG. 14A shows a masked likelihood function. FIG. 14B shows a regenerated likelihood function. The likelihood function regenerator 514 calculates the regenerated likelihood function l based on Equation (13) using the applied likelihood obtained by the processing of the applied likelihood recalculator 513, the applied orientation corresponding to the applied likelihood, and the orientation likelihood width parameter stored in the orientation likelihood width parameter storage unit 528. re,i t Generate (θ).

[0098]

number

[0099] The weighting unit 502 applies the regenerated likelihood function obtained by the processing of the applied likelihood recalculation unit 513 and the orientation change rate weight w obtained by the processing of the weight generation unit 510 to the t (θ') is used to perform weighting processing based on equation (14), and the weighted likelihood function L re_w,i t (θ,θ') is generated, where w t (θ') is a vector related to the rate of change of direction θ'. re,i t (θ)l re,i t (θ) is a vector related to the direction θ. The weighting unit 502 multiplies all combinations of θ and θ′ based on equation (14) and arranges the results as elements in a matrix.

[0100]

number

[0101] As described above, according to the tracking device 100 of the third embodiment, the maximum likelihood l extracted by the peak likelihood extracting unit 512 is peak The maximum likelihood buffer unit 527 stores a set amount of data. The applied likelihood recalculator 513 generates an applied likelihood. The likelihood function regenerator 514 generates a regenerated likelihood function based on the applied likelihood. Therefore, each tracker 500 generates a masked likelihood function l m,i t Instead of using the instantaneous likelihood (θ) to perform Bayesian estimation, we use the maximum past likelihood l peak The peak of the likelihood of the target to be tracked from the distribution is taken as the applied likelihood, and Bayesian estimation is performed using the regenerated likelihood function. As a result, in a situation where the target directions are close to each other, such as immediately after multiple targets cross each other as shown in Figure 7(b), the mask likelihood function l m,i t The maximum likelihood l due to multiple targets that cannot be distinguished and are combined in (θ) peak Therefore, for example, tracking unit 500-1 can generate a function of likelihood for target A, which is the original tracking target, as shown in FIG.

[0102] Fig. 15 is a diagram illustrating the effect of the tracking device 100 according to the third embodiment. In the tracking device 100 according to the third embodiment, as shown in Fig. 15, each tracking unit 500 can perform processing based on a likelihood function corresponding to the respective azimuth and azimuth change rate of the target to be tracked. As a result, it is possible to improve the tracking ability in tracking situations including when multiple targets intersect.

[0103] Embodiment 4 Fig. 16 is a diagram showing the configuration of tracking unit 500 according to embodiment 4. In Fig. 16, the components denoted by the same reference numerals as in Fig. 11 and the like perform the same processing as that explained in embodiments 1 to 3.

[0104] As shown in Fig. 16, the tracking unit 500 of the tracking device 100 in the fourth embodiment has an applied likelihood buffer unit 529 in the storage device 120 instead of the maximum likelihood buffer unit 527 in the third embodiment. The applied likelihood buffer unit 529 stores the applied likelihood obtained by the processing of the applied likelihood recalculation unit 513 for a set amount (time tT lapply T from time t lapply Here, the applied likelihood buffer unit 529 stores the maximum likelihood l extracted by the peak likelihood extraction unit 512 for a certain period of time after the start of target tracking. peak is stored as data, and after a certain time has passed, the applied likelihood is stored as data.

[0105] Next, the processing of the tracking unit 500 according to the third embodiment will be further described. Here, the processing of the applied likelihood recalculator 513 will be particularly described. The applied likelihood recalculator 513 calculates the maximum likelihood l for a set amount stored in the applied likelihood buffer 529. peak and the maximum likelihood l obtained by the processing of the peak likelihood extraction unit 512. peak and maximum orientation θ peak First, the applied likelihood recalculation unit 513 performs processing based on T lapply Maximum likelihood of l peak The average value μ lapply,i t Calculate.

[0106]

number

[0107] The applied likelihood recalculator 513 further calculates the maximum likelihood l peak and maximum orientation θ peak Using the average value μt, the application likelihood is calculated based on the above-mentioned equation (10) and the application direction is calculated based on the equation (11). Here, in the fourth embodiment, i peak.apply,i is expressed by equation (16).

[0108]

number

[0109] After a certain time has elapsed since the start of target tracking, the applied likelihood recalculator 513 stores the processed applied likelihood as data in the applied likelihood buffer 529 .

[0110] 17 is a diagram illustrating the effect of the tracking device 100 according to the fourth embodiment. First, as in the third embodiment, the maximum likelihood l peak In the case where the azimuths θ of a plurality of targets are close to each other, as shown in FIG. 17(a), the maximum likelihood buffer unit 527 of the tracking unit 500-2 that is performing the processing related to the target B also stores the maximum likelihood l related to the target A. peak,2 t As a result, if the two targets are in close proximity for a long time, after the two targets cross, the likelihood related to target A will be used in the processing of tracking unit 500-2, which may hinder accurate estimation of the direction of target B.

[0111] Therefore, tracking device 100 of embodiment 4 is configured to store the applied likelihood as data in applied likelihood buffer unit 529. Therefore, by storing the applied likelihood obtained by recalculation in applied likelihood recalculation unit 513 in applied likelihood buffer unit 529, the data used for calculation in applied likelihood recalculation unit 513 does not include data on likelihoods related to different targets. Therefore, as shown in FIG. 17(b), it is possible to prevent applied likelihood recalculation unit 513 from recalculating the applied likelihood based on the likelihood related to target A that is not being tracked by tracking unit 500-2. As a result, it is possible to further improve the tracking ability in tracking situations including when multiple targets intersect.

[0112] Embodiment 5. In the tracking device 100 according to the first and second embodiments described above, the azimuth change rate estimation unit 509 uses the prior probability distribution P prior,it+1|t(θ), but the processing is not limited to this. posterior Processing may be performed based on t(θ).

[0113] Regarding the weight generation unit 510 described in the first and second embodiments, equation (2) in the first embodiment and equation (8) in the second embodiment use a shape according to a Gaussian distribution, but this is not limitative. Any distribution that can be predicted in advance can be applied, not limited to a Gaussian distribution. In addition, the estimated direction θ stored in the estimated direction buffer unit 526 * Based on the orientation change rate obtained by the time difference of the above, the distribution of the orientation change rate can be estimated and applied.

[0114] In the third embodiment, the orientation likelihood width parameter of the orientation likelihood width parameter storage unit 528 is a parameter that is set in advance, but this is not limiting. For example, the maximum orientation θ obtained by the processing of the peak likelihood extraction unit 512 peak The azimuth likelihood width parameter may be set based on the variance of the azimuth likelihood width parameter.

[0115] In the above-described first to fourth embodiments, the data processing device 110 has two tracking units 500 and tracks two targets, but this is not limiting. By increasing the number of tracking units 500, it is possible to track three or more targets.

[0116] The tracking device 100 in the above-described first to fourth embodiments tracks a target based on the sound wave signals received by the microphones 210 of the microphone array 200. Here, the microphone array 200 is assumed to be mounted on a passive sonar, but is not limited to this. The present invention can also be applied to an active sonar that emits sound waves and converts the sound waves reflected from the target into acoustic signals. The present invention can also be applied to signals obtained by RADER (Radio Detecting and Ranging), LIDAR (Light Detection and Ranging), etc. [Explanation of symbols]

[0117] 100 Tracking device 110 Data processing device 120 Storage device 130 Signal input device 140 Signal output device 200 microphone array 210, 210-1 to 210-n microphones 220 Signal conversion device 230 Signal Transmitting Device 300 communication lines 400 Likelihood generation unit 500,500-1,500-2 Tracking part 501 Mask processing section 502 Weighting section 503 Bayesian Estimation Calculation Unit 504 Detection and judgment unit 505 ID Assignment Department 506 Orientation estimation part 507 Markov Update Department 508 Mask Generation Unit 509 Heading change rate estimation unit 510 Weight generation unit 511 Error standard deviation calculation section 512 Peak likelihood extraction unit 513 Applied Likelihood Recalculation Unit 514 Likelihood function regeneration unit 521 Prior probability buffer section 522 Mask memory section 523 Weight Memory Unit 524 Adjustment parameter memory 525 Estimated heading change rate buffer 526 Estimated direction buffer section 527 Maximum Likelihood Buffer 528 Direction likelihood width parameter storage unit 529 Applied Likelihood Buffer

Claims

1. a likelihood generation unit that calculates likelihoods in a plurality of directions from observation data obtained by observing signals related to a plurality of tracking targets and generates a likelihood function with the directions as parameter variables; a data processing device having a plurality of tracking units that determine the presence and direction of a tracking target corresponding to one of the plurality of tracking targets based on the likelihood function; Each of the tracking units comprises: an azimuth change rate estimation unit that performs an azimuth change rate estimation process to estimate an azimuth change rate in the corresponding azimuth of the tracking target and calculates an azimuth change rate estimated value; a weight generation unit that generates a heading change rate weight based on the heading change rate estimation value, and a weighting unit that generates a weighted likelihood function by superimposing the orientation change rate weight on the likelihood function; a Bayesian estimation calculation unit that performs processing to generate a posterior probability distribution from the weighted likelihood function; a Markov update unit that performs processing to generate a prior probability distribution; a detection determination unit that performs processing to determine the presence or absence of the tracking target based on the posterior probability distribution and detect the tracking target; an ID assigning unit that assigns an ID number to the tracking target detected by the detection and determination unit; an azimuth estimation unit that estimates the azimuth of the tracking target based on the posterior probability distribution and calculates an estimated azimuth; A tracking device having:

2. The tracking device according to claim 1 , wherein the azimuth change rate estimation unit performs the azimuth change rate estimation process based on the prior probability distribution generated by the Markov update unit.

3. The tracking device according to claim 1 , wherein the azimuth change rate estimation unit performs the azimuth change rate estimation process based on the posterior probability distribution generated by the Bayesian estimation calculation unit.

4. The tracking device according to any one of claims 1 to 3, wherein the weight generation unit generates the orientation change rate weight from data relating to a time series of the estimated orientation estimated by the orientation estimation unit and the orientation change rate estimated value estimated by the orientation change rate estimation unit.

5. Each of the tracking units comprises: a peak likelihood extracting unit that extracts a maximum likelihood from the likelihood function generated by the likelihood generating unit as a maximum likelihood; an applied likelihood recalculation unit that calculates an average of the maximum likelihoods for a preset period and calculates an applied likelihood and an applied orientation corresponding to the applied likelihood to be used in processing by the Bayesian estimation calculation unit; a likelihood function regeneration unit that regenerates the likelihood function based on the applied likelihood and the applied orientation; The tracking device according to any one of claims 1 to 4, comprising:

6. The tracking device according to claim 5, wherein the applied likelihood recalculation unit calculates an average of the applied likelihood for the set period after a certain time has elapsed since the start of tracking, and calculates the applied likelihood and the applied direction corresponding to the applied likelihood to be used in processing by the Bayesian estimation calculation unit.

7. a receiver array having a plurality of receivers and a signal converter that converts signals received by the receivers into digital data and generates observation value data; The tracking device according to any one of claims 1 to 6, A tracking system comprising:

8. A tracking method for tracking each of a plurality of tracking targets based on a likelihood function with a direction as a parameter variable, the likelihood function being generated by likelihoods related to a plurality of directions calculated from observation value data obtained by observing signals related to the plurality of tracking targets, the method comprising: a step of calculating an azimuth change rate estimated value by performing an azimuth change rate estimation process for estimating an azimuth change rate in the corresponding azimuth of the tracking target; generating a heading rate weight based on the heading rate estimate; and generating a weighted likelihood function by superimposing the heading change rate weight on the likelihood function; generating a posterior probability distribution from the weighted likelihood function; performing a process to generate a prior probability distribution; performing a process of determining the presence or absence of the tracking target based on the posterior probability distribution and detecting the tracking target; assigning an ID number to the detected tracking target; a step of estimating the direction of the tracking target based on the posterior probability distribution and calculating an estimated direction; A tracking method having the following.

9. A tracking method program for tracking each of a plurality of tracking targets based on a likelihood function with a direction as a parameter variable, the likelihood function being generated by likelihoods related to a plurality of directions calculated from observation value data obtained by observing signals related to the plurality of tracking targets, a step of calculating an azimuth change rate estimated value by performing an azimuth change rate estimation process for estimating an azimuth change rate in the corresponding azimuth of the tracking target; generating a heading rate weight based on the heading rate estimate; and generating a weighted likelihood function by superimposing the heading change rate weight on the likelihood function; generating a posterior probability distribution from the weighted likelihood function; performing a process to generate a prior probability distribution; performing a process of determining the presence or absence of the tracking target based on the posterior probability distribution and detecting the tracking target; assigning an ID number to the detected tracking target; a step of estimating the direction of the tracking target based on the posterior probability distribution and calculating an estimated direction; A tracking method program that allows a computer to perform the following.

Images