JP2009002799A - Apparatus, method, and program for target tracking - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reduce accuracy degradations in the positions of the center of gravity of detected positions and achieve highly accurate wake in a sensor system for acquiring a plurality of detected positions from the same target. <P>SOLUTION: A tracking processing part 108 tracks targets on the basis of detection information acquired from a detection device 200 through the use of an observation model in which an observation accuracy parameter indicating observation accuracy is set. A clustering computation part 104 computes a plot set indicating a set of positions recognized as targets indicated by the detection information acquired from the detection device 200. A size computation part 110 computes a plot set size, the size of the plot set. On the basis of the plot set size, a parameter control part 112 controls observation accuracy indicated by the observation accuracy parameter set in the observation model used by the tracking processing part 108. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a target tracking technique for tracking a target based on detection information obtained by a detection device such as a sensor, for example.

Observe the target with a high-resolution sensor. When the target is a large target, the high resolution sensor detects (observes) a plurality of detection positions from the same target. When performing target tracking with a high-resolution sensor, a tracking filter that performs target tracking generates a wake that is a target tracking result by using the barycentric positions of a plurality of detection positions.
JP 2004-251660 A

Depending on the positional relationship between the high resolution sensor and the target, the detection position may be lost. For this reason, there exists a subject that the precision of the gravity center position of a detection position deteriorates.
Further, when the position of the center of gravity of the detected position with degraded accuracy is input to the tracking filter, there is a problem that the accuracy of the wake that is the target tracking result is degraded.

An object of the present invention is to reduce deterioration in accuracy of the center of gravity of a detection position in a sensor system (target tracking system) that obtains a plurality of detection positions from the same target, for example.
Another object of the present invention is to improve the accuracy of a target track output from a tracking filter used in a sensor system.

The target tracking device according to the present invention is, for example, a target tracking device that acquires detection information indicating a position of a target from a detection device that detects a target and performs target tracking,
A detection information acquisition unit that acquires detection information from the detection device and stores the detection information in a storage device;
Based on the detection information acquired by the detection information acquisition unit using an observation model in which an observation accuracy parameter indicating observation accuracy is set, a tracking processing unit that tracks the target by the processing device;
Clustering a plot that is a target position indicated by the detection information acquired by the detection information acquisition unit, and a clustering calculation unit that calculates a plot set indicating a set of positions that are recognized as targets by the processing device;
A size calculation unit that calculates a plot set size, which is the size of the plot set calculated by the clustering calculation unit, by a processing device;
A parameter control unit for controlling the observation accuracy indicated by the observation accuracy parameter set in the observation model used by the tracking processing unit based on the size of the plot set size calculated by the size calculation unit by a processing device; And

  The target tracking device according to the present invention controls the observation accuracy indicated by the observation accuracy parameter set in the observation model based on the size of the plot set size. Therefore, the target tracking device according to the present invention can perform target tracking with high accuracy.

Embodiment 1 FIG.
In this embodiment, the target tracking device 100 that controls the observation accuracy indicated by the observation accuracy parameter set in the observation model based on the size of the plot set size will be described.

  First, the external appearance of the target tracking device 100 in this embodiment and the following embodiments will be described. FIG. 1 is a diagram illustrating an example of an appearance of a target tracking system including the target tracking device 100.

As shown in FIG. 1, the target tracking system includes, for example, a target tracking device 100 and a plurality of sensors 905. The target tracking device 100 includes an LCD 901 (Liquid Crystal Display), a keyboard 902 (K / B), a mouse 903, and a server 904.
The target tracking device 100 acquires detection information obtained by detecting the target aircraft 906 from each sensor 905. Then, the target tracking device 100 tracks the aircraft 906 based on the acquired detection information.

Next, the hardware configuration of the target tracking device 100 in this embodiment and the following embodiments will be described. FIG. 2 is a diagram illustrating an example of a hardware configuration of the target tracking device 100.
As shown in FIG. 2, the target tracking device 100 includes a CPU 911 (also referred to as a central processing unit, a central processing unit, a processing unit, an arithmetic unit, a microprocessor, a microcomputer, and a processor) that executes a program. The CPU 911 is connected to the ROM 913, the RAM 914, the LCD 901, the keyboard 902, the mouse 903, the communication board 915, and the magnetic disk device 920 via the bus 912, and controls these hardware devices.

  The ROM 913 and the magnetic disk device 920 are examples of a nonvolatile memory. The RAM 914 is an example of a volatile memory. The ROM 913 and the RAM 914 are examples of storage devices. The communication board 915 and the keyboard 902 are examples of input devices. The communication board 915 is an example of an output device. Furthermore, the communication board 915 is an example of a communication device. The LCD 901 is an example of a display device.

  An operating system 921 (OS), a window system 922, a program group 923, and a file group 924 are stored in the magnetic disk device 920 or the ROM 913. The programs in the program group 923 are executed by the CPU 911, the operating system 921, and the window system 922.

The program group includes the “detection information acquisition unit 102”, “clustering calculation unit 104”, “plot set centroid calculation unit 106”, “tracking processing unit 108”, “size calculation unit 110”, “parameter” in the following description. "Control unit 112", "display processing unit 114", "trajectory information storage unit 116", "gate control unit 118", "difference prediction information storage unit 120", "area determination processing unit 122", "difference determination unit 124" , A program for executing functions described as “difference control processing unit 126”, “plot monitoring unit 128”, “memory track control unit 130”, “shape storage unit 132”, “shape difference complementing unit 134”, and the like Is stored. The program is read and executed by the CPU 911.
In the file group, information, data, signal values, variable values, and parameters described as “detection information”, “control information”, etc. in the following description are stored as items of “file” and “database”. . The “file” and “database” are stored in a recording medium such as a disk or a memory. Information, data, signal values, variable values, and parameters stored in a storage medium such as a disk or memory are read out to the main memory or cache memory by the CPU 911 via a read / write circuit, and extracted, searched, referenced, compared, It is used for operations of the CPU 911 such as calculation / calculation / processing / output / printing / display. Information, data, signal values, variable values, and parameters are temporarily stored in the main memory, cache memory, and buffer memory during the operation of the CPU 911 for extraction, search, reference, comparison, calculation, calculation, processing, output, printing, and display. Is remembered.
In the following description, the arrows in the flowchart mainly indicate input / output of data and signals, and the data and signal values are recorded in a memory of the RAM 914 and other recording media such as an optical disk. Data and signals are transmitted online via a bus 912, signal lines, cables, or other transmission media.

  In addition, what is described as “to part” in the following description may be “to circuit”, “to device”, “to device”, “to means”, and “to function”. It may be “step”, “˜procedure”, “˜processing”. Further, what is described as “to process” may be “to step”. That is, what is described as “˜unit” may be realized by firmware stored in the ROM 913. Alternatively, it may be implemented only by software, only hardware such as elements, devices, substrates, wirings, etc., or a combination of software and hardware, and further a combination of firmware. Firmware and software are stored in a recording medium such as ROM 913 as a program. The program is read by the CPU 911 and executed by the CPU 911. That is, the program causes a computer or the like to function as “to part” described below. Alternatively, the procedure or method of “to part” described below is executed by a computer or the like.

Next, the function of the target tracking device 100 according to this embodiment will be described. FIG. 3 is a functional block diagram showing functions of the target tracking device 100 according to this embodiment.
The target tracking device 100 is connected to the detection device 200 via a network. The target tracking device 100 may be connected to a plurality of detection devices 200, but only one detection device 200 is illustrated here for simplicity.
The target tracking device 100 acquires detection information indicating the position of the target from the detection device 200 that detects the target, and performs target tracking.
The target tracking device 100 includes a detection information acquisition unit 102, a clustering calculation unit 104 (plot clustering processing unit), a plot set centroid calculation unit 106, a tracking processing unit 108 (tracking filter processing unit), and a size calculation unit 110 (plot set size monitoring). Section), a parameter control section 112 (filter parameter control section), and a display processing section 114.

The detection device 200 is, for example, a sensor such as a radar, IR (InfraRed), EW (Electronic Warfare), EO (Electronic Optical), or ESM (Electronic Support Measurements).
It is assumed that the detection apparatus 200 detects a target with high resolution. That is, the detection device 200 is a high resolution sensor.
There may be one detection device 200 or a plurality of detection devices. The detection device 200 may include a plurality of different types of sensors.

The detection information acquisition unit 102 acquires detection information indicating the target position (detection position) from the detection device 200 and stores the detection information in the storage device. The detection information acquired by the detection information acquisition unit 102 includes not only detection positions obtained from the same target but also detection positions and false alarms obtained from other targets.
The detection information acquisition unit 102 inputs the acquired detection information to the clustering calculation unit 104.

The detection position may be either a position indicated by one detection information detected by the detection apparatus 200 or a center of gravity of consecutive positions indicated by a plurality of detection information detected by the detection apparatus 200. .
FIG. 4 is a conceptual diagram of the position indicated by one detection information detected by the detection apparatus 200 and the center of gravity of successive positions indicated by a plurality of detection information. FIG. 4 is a conceptual diagram in a two-dimensional space when a plurality of detection positions are obtained from the same target when a large target is observed with a high resolution sensor, for example.
Among the cells, each colored cell is a detection position indicated by one detection information detected by the detection apparatus 200.
A set (A), B, and C of continuous (adjacent) detection positions are obtained from the large target. The centroids in the sets A, B, and C are the centroids A, B, and C. That is, the centroids A, B, and C are the centroids of consecutive positions detected by the detection device 200.
Note that the horizontal and vertical axes of the two-dimensional space in FIG. 4 may be the horizontal and vertical axes in various coordinate systems.

The clustering calculation unit 104 clusters plots that are target detection positions indicated by the detection information acquired by the detection information acquisition unit 102, and calculates a plot set that indicates a set of positions recognized as targets by the processing device.
For example, the clustering calculation unit 104 integrates continuous plots with a processing device.
Next, the clustering calculation unit 104 calculates the center of gravity of the integrated plot set by the processing device.
Then, the clustering calculation unit 104 uses the processing device to calculate a set of detection positions that can be regarded as a target, using the calculated one or plural centroids and the wake information generated by the tracking processing unit 108.
A set of detection positions that can be regarded as the target is a clustered detection position, which is called a plot set.

FIG. 5 is a diagram illustrating a conceptual diagram of clustered detection positions.
Each detection information acquired by the detection information acquisition unit 102 indicates each cell colored as a detection position.
The clustering calculation unit 104 integrates the continuous detection positions indicated by the detection information, and calculates the centroid of each continuous detection position. Then, the centroids at the three locations shown in FIG. 5 are calculated. The clustering calculation unit 104 calculates a set of positions indicated by broken lines as a plot set based on the calculated three centroids and track information generated by tracking in the past. That is, the clustering calculation unit 104 calculates that a set of positions indicated by broken lines in FIG. 5 is one target.

The clustering calculation unit 104 inputs the plot set and the detection information acquired by the detection information acquisition unit 102 to the plot set centroid calculation unit 106. Further, the clustering calculation unit 104 inputs the plot set size calculated by the size calculation unit 110 described later to the plot set centroid calculation unit 106.
Further, the clustering calculation unit 104 inputs the plot set and the detection information acquired by the detection information acquisition unit 102 to the size calculation unit 110.

The plot set centroid calculation unit 106 calculates the centroid of the set of clustered detection positions input by the clustering calculation unit 104.
FIG. 6 is a conceptual diagram of the center of gravity of the clustered detection positions. As shown in FIG. 6, the plot set centroid calculation unit 106 calculates the centroids of clustered detection positions indicated by broken lines.

The plot set centroid calculation unit 106 inputs the calculated clustered centroids of the detection positions and the detection information acquired by the detection information acquisition unit 102 to the tracking processing unit 108.
In addition, the plot set centroid calculation unit 106 inputs the plot set size calculated by the size calculation unit 110 described later to the tracking processing unit 108.

The tracking processing unit 108 inputs the centroids of the clustered detection positions calculated by the plot set centroid calculation unit 106 and the detection information acquired by the detection information acquisition unit 102 as observation values of the tracking filter, and the track information is processed by the processing device. Is generated.
Then, the tracking processing unit 108 inputs the generated track information to the display processing unit 114. In addition, the tracking processing unit 108 inputs the clustered detection positions, the detection information acquired by the detection information acquisition unit 102, and the plot set size to the display processing unit 114.

The tracking processing unit 108 sets a target motion model and an observation model by applying, for example, a Kalman filter, an α-β filter, an α-β-γ filter, an H infinity filter, or the like in control engineering as a tracking filter. To generate a target wake.
The track information includes, for example, time, track position, track speed (speed), track acceleration (speed change rate), observation value used to generate the track, target track number, sensor type used, memory track information, track The information includes the probability of the observed value, the management number of the detection position, the management number of the clustered detection position, the plot set size, and the like.
The memory track information is information on whether or not the tracking filter has extrapolated the wake of the previous time up to the current time by the motion model defined by the tracking filter without correlating with the target.

In the observation model set by the tracking filter, an observation accuracy parameter representing the ambiguity of the target observation is set.
The observation accuracy parameter indicates, for example, the accuracy of the polar coordinate distance, the angle accuracy, or the X, Y, and Z coordinates of the orthogonal coordinates.
In addition to the distance and angle accuracy of the target position, the observation accuracy parameter includes the distance corresponding to the target speed, the angle change rate, the X, Y, and Z speeds of Cartesian coordinates, and the speeds of various assumed coordinates. If the observation model can be set, etc., the accuracy may be set. Similarly, it is also possible to set the accuracy of the acceleration and jerk obtained by third-order and fourth-order differentiation of the target position, or the accuracy of specifications obtained by N-order differentiation of the target position.
Further, if the S / N ratio of the sensor can be observed and can be input to the tracking filter, the accuracy such as the S / N ratio may be set.
The accuracy indicated by the observation accuracy parameter may be set in any coordinate system.

Here, it is desirable that the observation accuracy parameter matches the accuracy of the observation value.
That the accuracy of the observation accuracy parameter and the observation value match means that the observation value input to the tracking filter matches the observation model set by the tracking filter. Therefore, as the observation accuracy parameter is closer to the accuracy of the observation value, the tracking filter assumes that the observation value is closer to the true value of the actual target, and the observation value is such that the contribution of the observation value to the track becomes larger. Perform wake correlation and observation filtering. Conversely, as the observation accuracy parameter is far from the accuracy of the observation, the tracking filter makes the observation different from the target true value, and the observation contributes less to the track. Performs correlation between values and wakes, and filtering of observation values.

In the motion model set by the tracking filter, a driving noise parameter representing the ambiguity of the target motion is set. The drive noise parameter is sometimes called a system noise parameter in control engineering.
When the driving noise parameter is 0, the target actual motion matches the motion model set in the tracking filter.

In the tracking filter, the wake is calculated by weighting the observed value and the predicted value internally calculated by the tracking filter.
The greater the drive noise parameter, the greater the ambiguity of the motion model set in the tracking filter. Therefore, the wake is calculated by performing weighting so that the wake contribution to the observed value becomes larger than the predicted value.
On the other hand, the smaller the drive noise parameter, the smaller the ambiguity of the motion model set by the tracking filter. Therefore, the wake is calculated by weighting so that the wake contribution to the predicted value is larger than the observed value.

  That is, the target tracking accuracy is improved by appropriately setting the observation accuracy parameter and the drive noise parameter. Therefore, the size calculation unit 110 and the parameter control unit 112, which will be described later, set appropriate observation accuracy parameters and drive noise parameters.

The size calculation unit 110 calculates the plot set size using the detection information input by the clustering calculation unit 104 and the plot set.
The size calculation unit 110 inputs the calculated plot set size to the parameter control unit 112.

A conceptual diagram of the plot set size is shown in FIGS.
FIG. 7 shows an area of clustered detection positions when the clustered detection positions are approximated by rectangles. That is, FIG. 7 is a conceptual diagram of a region for obtaining the plot set size when the clustered detection positions are approximated by rectangles. Approximating with a rectangle means, for example, obtaining a minimum rectangle including all detection positions.
FIG. 8 shows clustered detection position regions when approximated by an ellipsoid. That is, FIG. 8 is a conceptual diagram of a region for obtaining the plot set size when the clustered detection positions are approximated by an ellipsoid. Approximation with an ellipsoid means, for example, obtaining a minimum ellipsoid including all detection positions.
FIG. 9 shows clustered detection position regions when approximated by a polygon made up of end points of detection cells. That is, FIG. 9 is a conceptual diagram of a region for obtaining a plot set size in the case where clustered detection positions are approximated by a polygon composed of end points of detection cells. Approximation with a polygon made up of end points of detection cells means, for example, obtaining a polygon formed by connecting the end points of detection positions.
FIG. 10 shows clustered detection position regions when the target shape is approximated by a polygon. That is, FIG. 10 is a conceptual diagram of a region for obtaining the plot set size when the target shape is approximated by a polygon. Approximating the target shape with a polygon means, for example, obtaining a minimum shape similar to the target shape including all detection positions when the outline of the target shape is known to some extent.
As shown in FIG. 7 to FIG. 10, various shapes can be considered for the area for obtaining the plot set size, and any shape may be used. Further, the area for obtaining the plot set size may be an N-dimensional space. N is an arbitrary natural number.

The size calculation unit 110 calculates the plot set size based on the regions shown in FIGS. 7 to 10 and other possible region shapes.
The plot set size is, for example, the area of the detection cell, the minimum value and the maximum value of the end points of the region when the plot set size is obtained.

FIG. 11 is a conceptual diagram illustrating an example of a method for calculating the plot set size.
FIG. 11 shows, as an example, a plot set size calculation method when the clustered detection positions shown in FIG. 7 are approximated by rectangles.
In FIG. 11, for example, the horizontal axis is X and the vertical axis is Y. Here, X and Y may be interchanged. X may be a range and Y may be a cross range. Further, various criteria such as X as a distance and Y as a direction may be used.
The end points on the horizontal axis and the vertical axis of the rectangular area are represented as X _max , Y _max , X _min , and Y _min , respectively.

  For example, the plot set size is calculated as in Equation 1 or Equation 2. The use of either Equation 1 or Equation 2 is predetermined.

Ω (k) = max [X _max −X _min , Y _max −Y _min ] (Formula 1)

Ω (k) on the left side of Equation 1 is the plot set size at time k.
The right side of Equation 1 means that the larger one of the lengths of the vertical and horizontal axes of the region is selected.

Ω (k) = (X _max −X _min ) × (Y _max −Y _min ) (Formula 2)

Ω (k) on the left side of Equation 2 is the plot set size at time k.
The right side of Equation 2 is obtained by multiplying the lengths of the vertical and horizontal axes of the region, and means the area.

Here, Formula 1 and Formula 2 are considered two-dimensional, but may be three-dimensional or N-dimensional. N is an arbitrary natural number.
In addition, when Formula 2 is considered in three dimensions, it means a volume.

The parameter control unit 112 controls the observation accuracy indicated by the observation accuracy parameter set in the observation model used by the tracking processing unit 108 based on the plot set size input by the size calculation unit 110 by the processing device.
The parameter control unit 112 controls the observation accuracy indicated by the observation accuracy parameter based on Equation 3 and Equation 4, for example.

| Ω (k) −Ω _g |> Ω _th (Formula 3)

| Ω (k) −Ω _g | ≦ Ω _th (Formula 4)

Ω _g in Equation 3 and Equation 4 is an average of plot set sizes for the past N times. Further, _Ωth is a threshold value (third threshold value) of the plot set size.
It should be noted, Ω in the calculation of the _g, take the average of the plot set the size of the N time worth, including the current time k, do not include the current time k, it may take the average of the plot set the size of the N time minutes. Method of calculating the Ω _g is, is determined in advance in advance.
Further, _Ωth is set to have a different value between the case where the current time plot set size is calculated using Equation 1 and the case where the current time plot set size is calculated using Equation 2.

When Expression 3 is satisfied, the parameter control unit 112 inputs an observation accuracy parameter whose observation accuracy is lowered to the tracking processing unit 108. Decreasing the observation accuracy means increasing the observation accuracy parameter.
On the other hand, when Expression 4 is satisfied, the parameter control unit 112 inputs an observation accuracy parameter with higher observation accuracy to the tracking processing unit 108. Increasing the observation accuracy means reducing the observation accuracy parameter.
Increasing (decreasing) the observation parameter means increasing (decreasing) the predetermined observation parameter that is normally used.
Further, the parameter control unit 112 changes the value of the driving noise parameter as necessary according to the plot set size and inputs the value to the tracking processing unit 108.

  The reason why the parameter control unit 112 performs such control is that the plot set size is significantly different from the past value because there is a problem with the observed value. That is, when the plot set size is significantly different from the past value, the parameter control unit 112 determines that there is a problem with the observed value and increases the influence of the predicted value based on the past observed value on the wake.

  The display processing unit 114 processes the track information input by the tracking processing unit 108, the clustered detection positions, and the detection positions so that the operator can easily see them, and displays them on the display device.

As described above, the target tracking device 100 according to this embodiment obtains a plot set size, and a plot determined in advance has an absolute value of a difference between the plot set size at the current time and the plot set size for the past N times. When the set size is larger than the threshold value, the observation accuracy parameter given to the tracking filter is set to be large. As a result, the tracking filter generates a wake by performing control so that the past tracking result is more important than the detection position at the current time.
Therefore, deterioration of wake accuracy is reduced. Furthermore, the center of gravity of the clustered detection positions can be calculated with high accuracy.

Embodiment 2. FIG.
In this embodiment, a target tracking device 100 that controls gates that narrow down detection information used for target tracking will be described.

First, the function of the target tracking device 100 according to this embodiment will be described. FIG. 12 is a functional block diagram showing functions of the target tracking device 100 according to this embodiment.
In addition to the function of the target tracking device 100 according to the first embodiment, the target tracking device 100 according to this embodiment further includes a trajectory information storage unit 116 (track information database) and a gate control unit 118.
Other functions of the target tracking device 100 according to this embodiment are substantially the same as the functions of the target tracking device 100 according to the first embodiment. Therefore, only the parts different from the target tracking device 100 according to Embodiment 1 will be described here.

  The trajectory information storage unit 116 stores, in a storage device, track information indicating the result of target tracking generated by the tracking processing unit 108 following the target.

The gate control unit 118 acquires the track information for the past predetermined period (for the past several hours) stored in the trajectory information storage unit 116.
Next, the gate control unit 118 acquires the number of times of memory tracking from the memory track information included in the acquired track information.
In addition, the gate control unit 118 acquires the plot set size included in the wake information. Furthermore, the gate control unit 118 calculates time-series change information indicating a change in the plot set size in the past predetermined period from the acquired plot set size. In addition, when the time series change information of the plot set size is included in the wake information, the time series change information may be acquired from the wake information.
Then, the gate control unit 118 limits the gate in the tracking filter based on the number of memory tracks and the time series change information of the plot set size. Specifically, the gate control unit 118 detects the detection information used by the tracking processing unit 108 for tracking when the number of times of memory tracking is a predetermined number or more, or when the plot set size is larger than a predetermined number. The processing device is limited so as not to widen the gate for narrowing down.
The plot set size being larger than a predetermined value may be either that the plot set size is larger than a predetermined size or that the plot set size is larger than a predetermined ratio.
Further, the gate control unit 118 may limit the gate based only on either the number of memory tracks or the rate of change of the plot set size.

FIG. 13 is a conceptual diagram of a gate in the tracking filter.
The gate narrows down detection information candidates used for tracking from the correspondence between the predicted value of the wake and the detection information (observed value). For example, the gate is formed of an ellipse centered on the predicted value as shown in FIG. 13, and detection information indicating a position included in the range of the ellipse is used for tracking. The shape of the gate is not limited to an ellipse, and may be any shape such as a circle, a rectangle, or a fan.
The gate may be a variable gate whose size varies depending on the wake and the error covariance matrix of observation values, or may be a fixed gate whose size is fixed.

The reason why the gate control is performed based on the number of memory tracks and the plot set size is as follows.
For example, when an observation value from another target existing in another adjacent area enters the target gate, the centroids of the clustered detection positions are shifted. As a result, a memory track is generated. In other words, if there are many memory tracks, the observations from another target may have entered the target gate many times.
Also, assuming that the gate is a variable gate, the gate may become large due to the occurrence of a memory track. As a result, the correlation with the detection information obtained from another target is obtained, and as a result, the centers of gravity of the clustered detection positions are shifted.
Furthermore, when the plot set size becomes large, there is a possibility that an observation value from another target existing in another adjacent area has entered the target gate.
Therefore, the target tracking device 100 according to this embodiment prevents a cross-correlation with detection information obtained from another target by limiting the gate based on the number of memory tracks and the change rate of the plot set size.

  As described above, the target tracking device 100 according to this embodiment calculates the centroids of clustered detection positions with high accuracy in order to prevent cross-correlation with detection information obtained from another target by limiting gates. be able to.

Embodiment 3 FIG.
In this embodiment, a description will be given of the target tracking device 100 that controls the observation accuracy parameter on the basis of the possibility of detection deviation or omission by the detection device 200.

First, the function of the target tracking device 100 according to this embodiment will be described. FIG. 14 is a functional block diagram showing functions of the target tracking device 100 according to this embodiment.
In addition to the function of the target tracking device 100 according to the first embodiment, the target tracking device 100 according to this embodiment further includes a difference prediction information storage unit 120 (area information database) and an area determination processing unit 122.
Other functions of the target tracking device 100 according to this embodiment are substantially the same as the functions of the target tracking device 100 according to the first embodiment. Therefore, only the parts different from the target tracking device 100 according to Embodiment 1 will be described here.

The difference prediction information storage unit 120 stores the difference between the detection information that should be obtained by the detection device 200 and the detection information that is actually expected to be obtained by the detection device 200 as the prediction difference information in the storage device for each predetermined area. Remember.
That is, the difference prediction information storage unit 120 stores information about areas where clustered detection positions are shifted or missing areas. The difference prediction information storage unit 120 indicates, for each predetermined area, detection position difference information indicating how much the detection position is shifted when the target is observed with a high resolution sensor, and how much the detection position is missing. The detected position missing information is stored as expected difference information. That is, the expected difference information includes detected position difference information, detected position missing information, and position information.

  The area determination processing unit 122 includes a difference determination unit 124 and a difference control processing unit 126.

The difference determination unit 124 acquires track information from the tracking processing unit 108.
The difference determination unit 124 identifies in which area the tracked target currently exists using the track information. That is, the difference determination unit 124 identifies which area is the position indicated by the latest detection information acquired by the detection information acquisition unit 102.
The difference determination unit 124 acquires predicted difference information of an area including the identified position (position where the target is currently) from the predicted difference information stored in the difference prediction information storage unit 120.
The difference determination unit 124 determines whether or not there is a difference in the acquired area by the processing device.

When the difference determination unit 124 determines that there is a difference, the difference control processing unit 126 inputs control information to the parameter control unit 112 by the processing device so that the observation accuracy indicated by the observation accuracy parameter set in the observation model is lowered. And give instructions.
On the other hand, if the difference determination unit 124 determines that there is no difference, the difference control processing unit 126 controls the parameter control unit 112 by the processing device so that the observation accuracy indicated by the observation accuracy parameter set in the observation model is increased. Enter information and give instructions.

  Here, it is determined whether priority is given to the setting of the magnitude of the observation accuracy parameter based on the plot set size described in the first embodiment, or whether the track of the tracking target described in the present embodiment is present in the missing area. It is determined in advance whether to prioritize the setting of the observation accuracy parameter size using. The observation accuracy parameter may be set in consideration of both.

As described above, the target tracking device 100 according to this embodiment pre-registers an area where clustered detection positions are expected to be missing, and sets a large observation accuracy parameter in the tracking filter.
Therefore, according to the target tracking device 100 according to this embodiment, it is possible to reduce deterioration in accuracy of the center of gravity of the clustered detection positions.

Embodiment 4 FIG.
In this embodiment, a target tracking device 100 that generates a memory track in a predetermined case will be described.

First, the function of the target tracking device 100 according to this embodiment will be described. FIG. 15 is a functional block diagram showing functions of the target tracking device 100 according to this embodiment.
The target tracking device 100 according to this embodiment includes a plot monitoring unit 128 and a memory track control unit 130 in addition to the functions of the target tracking device 100 according to the first embodiment.
Other functions of the target tracking device 100 according to this embodiment are substantially the same as the functions of the target tracking device 100 according to the first embodiment. Therefore, only the parts different from the target tracking device 100 according to Embodiment 1 will be described here.

The plot monitoring unit 128 acquires the plot set size and the number of plots from the size calculation unit 110.
The number of plots is the number of plots included in the plot set calculated by the clustering calculation unit 104. The plot monitoring unit 128 acquires the number of plots from the clustering calculation unit 104 via the size calculation unit 110.
Next, the plot monitoring unit 128 determines based on the plot set size whether the target plot set size is larger than a predetermined threshold.
Further, the plot monitoring unit 128 determines whether the number of plots is greater than a predetermined threshold based on the number of plots.
Then, the plot monitoring unit 128 inputs the determined result to the memory track control unit 130.

Based on the determination result of the plot monitoring unit 128, the memory track control unit 130 determines that the plot set size is larger than a predetermined value (first threshold) or the number of plots included in the plot set is a predetermined value (first number). If it is greater than the threshold (2), the control device transmits control information to the tracking processing unit 108 so as to generate a memory track.
That is, in the above case, the memory track control unit 130 determines that the detection information is invalid, and sets the tracking processing unit 108 so as to perform tracking of the target without being based on the detection information that is the observation value at the current time. Control. By controlling in this way, the track is forcibly prevented from leaving a predetermined range.

As described above, the target tracking device 100 according to this embodiment forcibly stores the tracking target so that the wake does not go out of the predetermined range when the plot set size or the number of plots becomes extremely large. Let it track. That is, the target tracking device 100 according to this embodiment has many false alarms (false detections) due to, for example, weather and building shielding information, the state of proximity to the target, the ground, the sea level, clouds, and the like. In this case, the tracking target is forcibly memory-tracked so that the wake does not go out of the predetermined range because the detection is invalid.
Therefore, according to the target tracking device 100 according to this embodiment, it is possible to reduce deterioration in accuracy of the center of gravity of clustered detection positions.

Embodiment 5. FIG.
In the fourth embodiment, the wake is controlled so as not to leave a predetermined range by generating a memory track in a predetermined case. In this embodiment, the wake does not leave a predetermined range by controlling the observation accuracy parameter in the predetermined case.

First, the function of the target tracking device 100 according to this embodiment will be described. FIG. 16 is a functional block diagram showing functions of the target tracking device 100 according to this embodiment.
The target tracking device 100 according to this embodiment has a configuration similar to that of the target tracking device 100 according to the fourth embodiment. Therefore, only the parts different from the target tracking device 100 according to the fourth embodiment will be described here.

Based on the determination result of the plot monitoring unit 128, the memory track control unit 130 determines that the plot set size is larger than a predetermined value (first threshold) or the number of plots included in the plot set is a predetermined value (first number). If it is greater than the threshold value (2), control information for controlling the observation accuracy indicated by the observation accuracy parameter to be set low is transmitted to the parameter control unit 112 by the processing device.
That is, in the above case, the memory track control unit 130 sets the observation accuracy parameter to a large value, assuming that the detection information is invalid. By controlling in this way, the track is forcibly prevented from leaving a predetermined range.

The parameter control unit 112 inputs a larger observation accuracy parameter to the tracking processing unit 108 based on the control information transmitted by the memory track control unit 130.
In addition, the parameter control unit 112 may input to the tracking processing unit 108 control information for performing a memory track so that the wake does not leave a predetermined range.

As described above, the target tracking device 100 according to this embodiment sets the observation accuracy parameter large so that the wake does not go out of the predetermined range when the plot set size or the number of plots becomes extremely large. . That is, the target tracking device 100 according to this embodiment has many false alarms (false detections) due to, for example, weather and building shielding information, the state of proximity to the target, the ground, the sea level, clouds, and the like. In this case, the observation accuracy parameter is set to be large so that the wake does not go out of the predetermined range because the detection is inappropriate. That is, an effect that is almost equivalent to forcing the tracking target to the memory track is obtained by setting the observation accuracy parameter large.
Therefore, according to the target tracking device 100 according to this embodiment, it is possible to reduce the accuracy deterioration of the center of gravity of the clustered detection positions.

Embodiment 6 FIG.
In this embodiment, a description will be given of the target tracking device 100 that complements data when the detection information obtained by the detection device 200 is missing.

First, the function of the target tracking device 100 according to this embodiment will be described. FIG. 17 is a functional block diagram showing functions of the target tracking device 100 according to this embodiment.
In addition to the function of the target tracking device 100 according to the first embodiment, the target tracking device 100 according to this embodiment further includes a trajectory information storage unit 116, a shape storage unit 132 (plot set information database), and shape difference complementation. Unit 134 (missing position calculation unit).
Other functions of the target tracking device 100 according to this embodiment are substantially the same as the functions of the target tracking device 100 according to the first embodiment. Therefore, only portions related to the trajectory information storage unit 116, the shape storage unit 132, and the shape difference complementing unit 134 will be described here.

  The trajectory information storage unit 116 is the same as that of the second embodiment, and stores the track information indicating the result of target tracking generated by the tracking processing unit 108 by tracking the target in the storage device.

The shape storage unit 132 acquires the clustered detection positions and the centroids of the clustered detection positions from the plot set centroid calculation unit 106 and stores them in the storage device. The centroid of the clustered detection position at the current time may be the centroid of the clustered detection position before update (one hour before).
In addition, the shape storage unit 132 acquires the plot set size from the size calculation unit 110 and stores it in the storage device.
And the shape memory | storage part 132 calculates a target shape based on the clustered detection position, the gravity center of the clustered detection position, and the plot set size, and memorize | stores it in a memory | storage device.

The shape difference complementing unit 134 acquires plot set information from the past to the present including the clustered detection positions, the centroids of the clustered detection positions, and the plot set size from the shape storage unit 132.
In addition, the shape difference complementing unit 134 acquires the target shape from the past to the present from the shape storage unit 132.
Further, the shape difference complementing unit 134 acquires track information from the past to the present.

The shape difference complementing unit 134 matches the target shape and plot set size included in the past plot set information with the target shape and plot set size included in the current plot set information.
Further, the shape difference complementing unit 134 matches a track obtained from past track information with a track obtained from current track information.
Then, the shape difference complementing unit 134 calculates the missing position of the detection position indicated by the detection information used to generate the current track information.
In other words, the shape difference complementation unit 134 compares the target shape of the past predetermined period with the latest target shape to obtain the difference, thereby detecting the latest target shape. The missing position that is missing in the information is calculated by the processing device, and missing information that complements the missing position is generated.

If there are many missing positions as a result of the calculation, the shape difference complementing section 134 inputs the missing information to the clustering calculating section 104, and again inputs control information for clustering the detected positions and calculating the plot set. .
The clustering calculation unit 104 calculates a plot set again based on the detection information acquired by the detection information acquisition unit 102 and the missing information input by the shape difference complementing unit 134.
The clustering calculation unit 104 inputs a plot set considering the newly calculated missing information to the plot set centroid calculation unit 106.

The shape difference complementing unit 134 inputs the missing information to the plot set centroid calculating unit 106.
The plot set centroid calculating unit 106 obtains the centroid of the plot set considering the missing information based on the missing information input by the shape difference complementing unit 134 and the plot set considering the missing information input by the clustering calculating unit 104. .
The plot set center-of-gravity calculation unit 106 inputs the plot set considering the missing information and the center of the plot set considering the missing information to the tracking processing unit 108.

  The tracking processing unit 108 performs target tracking again based on the plot set considering missing information and the centroid of the plot set considering missing information.

As described above, the target tracking device 100 according to this embodiment complements the missing clustered detection positions at the current time based on the clustered detection positions and target shapes observed in the past.
Therefore, according to the target tracking device 100 according to this embodiment, it is possible to reduce the accuracy deterioration of the center of gravity of the clustered detection positions.

That is, the target tracking device 100 according to the above embodiment is
A detection device 200 (high resolution sensor) that obtains a plurality of detection positions from the same target;
A set of continuous detection positions is generated from the detection positions input from the detection device 200, a centroid consisting of the continuous detection positions is obtained for each continuous detection position, and the detection positions are determined using these centroids as detection positions. Clustering calculation unit 104 (plot clustering processing unit) for obtaining clustered detection positions;
A plot set center-of-gravity calculation unit 106 for obtaining the center of gravity of the clustered detection positions from the clustered detection positions;
The tracking processing unit 108 (tracking filter processing unit) that obtains track information by the tracking filter using the clustered detection position centroid as the observed value of the tracking filter;
A display processing unit 114 for displaying the wake information in an easy-to-see manner for the operator;
A size calculation unit 110 (plot set size) that approximates an area of clustered detection positions using a clustered detection position and calculates a plot set size based on the length or area of the shape and an N-dimensional volume. Monitoring department),
The observation accuracy parameter set for the tracking filter based on whether the absolute value of the difference between the plot set size for the past several hours and the plot set size at the current time exceeds the preset plot set size threshold And a parameter control unit 112 (filter parameter control unit) for controlling the driving noise parameter;
It is provided with.

Furthermore, the target tracking device 100 according to the above embodiment includes:
A track information storage unit 116 (track information database) for storing track information;
A gate control unit 118 that limits the size of the tracking filter gate by looking at the time-series changes in the memory track information and the plot set size from the track information for the past several hours,
It is provided with.

Furthermore, the target tracking device 100 according to the above embodiment includes:
The difference prediction information storage unit 120 (area information database) that accumulates area information in which the detection position missing information includes the position of the area;
Using the track and area position calculated by the tracking filter, a signal for controlling the observation accuracy parameter is output to the filter parameter control unit depending on whether or not it is determined that a track exists in the area where the detection position is missing. An area determination processing unit 122 for controlling the observation accuracy parameter of the filter;
It is provided with.

Furthermore, the target tracking device 100 according to the above embodiment includes:
From the plot set size and the number of plots information, whether the plot set size is larger than the predetermined plot set size threshold, or the number of plots is larger than the predetermined plot number threshold A plot monitoring unit 128 that outputs information on whether or not the value is large;
From the information on whether the plot set size is larger than the predetermined plot set size threshold or the information on whether the number of plots is larger than the predetermined plot number threshold, A memory track control unit 130 for outputting a control signal for memory tracking so as not to use the observation value at the current time;
It is provided with.

Furthermore, the memory track control unit 130
From the information on whether the plot set size is larger than the predetermined plot set size threshold or the information on whether the number of plots is larger than the predetermined plot number threshold, In order not to use the observation value at the current time, the control for equivalent memory tracking is performed by setting the observation accuracy parameter set in the tracking filter to a large value.

Furthermore, the target tracking device 100 according to the above embodiment includes:
A shape storage unit 132 (plot set information database) for accumulating plot set information including clustered detection positions, centroids of clustered detection positions, and target shapes calculated in advance;
A shape difference complementing unit 134 (missing position) that calculates an estimated missing position from the plot set information and the wake information and outputs a control signal that takes the center of gravity for the interpolated estimated missing position at the clustered detection position. Calculation part),
It is provided with.

The figure which shows an example of the external appearance of a target tracking system provided with the target tracking apparatus. The figure which shows an example of the hardware structure of the target tracking apparatus. FIG. 3 is a functional block diagram illustrating functions of the target tracking device 100 according to the first embodiment. The conceptual diagram of the gravity center of the position which one detection information which the detection apparatus 200 detected, and the continuous position which several detection information shows. The figure which shows the conceptual diagram of the clustered detection position. The conceptual diagram of the gravity center of the clustered detection position. The conceptual diagram of the area | region when calculating | requiring plot set size when the clustered detection position is approximated by the rectangle. The conceptual diagram of the area | region when calculating | requiring plot set size at the time of approximating the clustered detection position with an ellipsoid. The conceptual diagram of the area | region when calculating | requiring the plot set size at the time of approximating the clustered detection position with the polygon which consists of the end point of a detection cell. The conceptual diagram of the area | region when calculating | requiring plot set size when a target shape is approximated with a polygon. The conceptual diagram which shows an example of the calculation method of plot set size. FIG. 4 is a functional block diagram illustrating functions of a target tracking device 100 according to a second embodiment. The conceptual diagram of the gate in a tracking filter. FIG. 9 is a functional block diagram showing functions of a target tracking device 100 according to a third embodiment. FIG. 9 is a functional block diagram showing functions of a target tracking device 100 according to a fourth embodiment. FIG. 6 is a functional block diagram showing functions of a target tracking device 100 according to a fifth embodiment. FIG. 10 is a functional block diagram showing functions of a target tracking device 100 according to a sixth embodiment.

Explanation of symbols

  100 target tracking device, 102 detection information acquisition unit, 104 clustering calculation unit, 106 plot set centroid calculation unit, 108 tracking processing unit, 110 size calculation unit, 112 parameter control unit, 114 display processing unit, 116 locus information storage unit, 118 Gate control unit, 120 Difference prediction information storage unit, 122 Area determination processing unit, 124 Difference determination unit, 126 Difference control processing unit, 128 Plot monitoring unit, 130 Memory track control unit, 132 Shape storage unit, 134 Shape difference complement unit 901 LCD, 902 keyboard, 903 mouse, 904 server, 911 CPU, 912 bus, 913 ROM, 914 RAM, 915 communication board, 920 magnetic disk unit, 921 operating system, 922 window system, 923 Program group, 924 file group.

Claims (10)

A target tracking device that acquires detection information indicating the position of the target from a detection device that detects the target and tracks the target,
A detection information acquisition unit that acquires detection information from the detection device and stores the detection information in a storage device;
Based on the detection information acquired by the detection information acquisition unit using an observation model in which an observation accuracy parameter indicating observation accuracy is set, a tracking processing unit that tracks the target by the processing device;
Clustering a plot that is a target position indicated by the detection information acquired by the detection information acquisition unit, and a clustering calculation unit that calculates a plot set indicating a set of positions that are recognized as targets by the processing device;
A size calculation unit that calculates a plot set size, which is the size of the plot set calculated by the clustering calculation unit, by a processing device;
A parameter control unit for controlling the observation accuracy indicated by the observation accuracy parameter set in the observation model used by the tracking processing unit based on the size of the plot set size calculated by the size calculation unit by a processing device; Target tracking device. The parameter control unit sets an observation accuracy indicated by an observation accuracy parameter set in an observation model used by the tracking processing unit to be low when the plot set size is larger than a first threshold. The target tracking device according to 1. The parameter control unit sets the observation accuracy indicated by the observation accuracy parameter set in the observation model used by the tracking processing unit when the number of plots included in the plot set is greater than a second threshold. The target tracking device according to claim 1 or 2. The size calculation unit calculates an average value of the target plot set size calculated by the size calculation unit during a predetermined period in the past by the processing device,
When the absolute value of the difference between the plot set size and the average value of the plot set size is greater than a third threshold, the parameter control unit is configured to perform an observation indicated by an observation accuracy parameter set in the observation model used by the tracking processing unit. The target tracking device according to any one of claims 1 to 3, wherein the accuracy is set low. The tracking processing unit tracks the target based on the detection information and a predicted position where the target position is predicted by a predetermined method.
The target tracking device further includes:
When the plot set size is larger than the first threshold or when the number of plots included in the plot set is larger than the second threshold, a memory track for tracking the target is generated without being based on the detection information. The target tracking device according to any one of claims 1 to 4, further comprising a memory track control unit that transmits control information to the tracking processing unit by a processing device. The target tracking device further includes:
A trajectory information storage unit that stores trajectory information indicating a result of the tracking processing unit tracking the target in a storage device;
Acquiring at least one of the number of times of memory tracking from the target trajectory information in the past predetermined period from the trajectory information stored in the trajectory information storage unit and the change information of the size of the plot set in the past predetermined period; When the number of times tracked is a predetermined number or more, or when the change information of the size of the plot set indicates that the plot set size is larger than the predetermined number, the detection information used by the tracking processing unit for tracking The target tracking device according to claim 1, further comprising a gate control unit that restricts by a processing device so as not to widen a gate for narrowing down. The target tracking device further includes:
A difference prediction information storage unit that stores a difference between detection information to be obtained by the detection device and detection information that is actually expected to be obtained by the detection device as prediction difference information in a storage device for each predetermined area;
The prediction difference information of the area including the position indicated by the detection information acquired by the detection information acquisition unit is acquired from the prediction difference information stored in the difference prediction information storage unit, and whether or not there is a difference in the area is processed. A difference determination unit to be determined by the device;
A difference control processing unit for instructing the parameter control unit by a processing device to lower the observation accuracy indicated by the observation accuracy parameter set in the observation model when the difference determination unit determines that there is a difference. The target tracking device according to any one of claims 1 to 6, wherein The target tracking device further includes:
A shape storage unit for storing a target shape in a storage device;
The latest shape of the target is calculated by comparing the target shape of the past predetermined period stored in the shape storage unit with the latest shape of the target calculated by the shape calculation unit to obtain a difference. A shape difference complementing unit that complements the detection information used for the processing by a processing device,
The target tracking apparatus according to claim 1, wherein the tracking processing unit performs target tracking again based on the detection information supplemented by the detection information complementing unit. It is a target tracking method of a target tracking device that acquires detection information indicating a target position from a detection device that detects a target and tracks the target,
A detection information acquisition step in which the storage device acquires and stores the detection information from the detection device; and
A processing device performs tracking of a target based on the detection information acquired in the detection information acquisition step using an observation model in which an observation accuracy parameter indicating observation accuracy is set; and
A clustering calculation step for clustering a plot which is a target position indicated by the detection information acquired in the detection information acquisition step, and a plot set indicating a set of positions recognized as targets;
A size calculation step in which the processing device calculates a plot set size that is a size of the plot set calculated in the clustering calculation step;
And a parameter control step for controlling the observation accuracy indicated by the observation accuracy parameter set in the observation model used in the tracking processing step based on the size of the plot set size calculated in the size calculation step. Feature target tracking method. It is a target tracking program that acquires detection information indicating the position of the target from a detection device that detects the target and tracks the target,
Detection information acquisition processing for acquiring detection information from the detection device and storing it in the storage device;
Using the observation model in which the observation accuracy parameter indicating the observation accuracy is set, based on the detection information acquired in the detection information acquisition processing, tracking processing processing for tracking the target by the processing device;
Clustering a plot that is a target position indicated by the detection information acquired in the detection information acquisition process, and a clustering calculation process that calculates a plot set that indicates a set of positions that are recognized as targets by a processing device;
A size calculation process for calculating a plot set size, which is the size of the plot set calculated in the clustering calculation process, by a processing device;
Based on the size of the plot set size calculated in the size calculation process, the computer executes a parameter control process for controlling the observation accuracy indicated by the observation accuracy parameter set in the observation model used in the tracking process process by the processing device. A goal tracking program characterized by that.

Images