Target image processing device and target image processing program

Abstract

This disclosure aims to detect small targets not only by detecting large targets but also by detecting small targets using a radar device, by clustering cells in which the radar received power exceeds a "small received power threshold," and to detect small targets separately from system noise even when the small targets are in close proximity to system noise. [Solution] The Doppler velocity variance of the target is smaller than that of the system noise. Therefore, in this disclosure, the Doppler velocity variance of each target cell within each cluster is calculated, and when the Doppler velocity variance of each target cell within each cluster is greater than or equal to the Doppler velocity variance threshold, each target cell within each cluster is separated from the cluster.

Description

Technical Field

[0001] The present disclosure relates to a technique for separating and detecting a target from system noise using a radar device.

Background Art

[0002] Techniques for separating and detecting targets such as ships from clutter such as waves using a radar device are disclosed in Patent Document 1 and the like. In Patent Document 1, in a target image acquired from a radar device, cells in which the radar reception power exceeds a reception power threshold value are clustered using DBSCAN (Density-Based Spatial Clustering of Applications with Noise) or the like, and clusters are extracted.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The problems to be solved by the conventional cluster extraction process are shown in FIG. 1. In Patent Document 1, in order to detect not only targets such as large ships but also targets such as small ships V or buoys, it is necessary to cluster cells in which the radar reception power exceeds a "small reception power threshold value". However, when a target such as a small ship V or a buoy is close to the system noise N, the target such as a small ship V or a buoy cannot be separated from the system noise N and detected.

[0005] Then, when tracking a target such as a small ship V or a buoy, the target such as a small ship V or a buoy is integrated with the system noise N, and the feature amounts (center of gravity position, irradiated area, maximum reception power, average Doppler speed, etc.) of the tracked target are shifted, resulting in low accuracy of target tracking.

[0006] Therefore, in order to solve the aforementioned problems, this disclosure aims to detect small targets not only by detecting large targets but also by detecting small targets using a radar device, by clustering cells in which the radar received power exceeds a "small received power threshold", and to detect small targets separately from system noise even when the small targets are in close proximity to system noise. [Means for solving the problem]

[0007] The Doppler velocity variance of the target is smaller than that of the system noise. To solve the above problem, the Doppler velocity variance of each target cell within each cluster is calculated, and when the Doppler velocity variance of each target cell within each cluster is greater than or equal to the Doppler velocity variance threshold, each target cell within each cluster is separated from the cluster.

[0008] Specifically, this disclosure relates to a target image acquisition unit that acquires a target image from a radar device, In the aforementioned target image, a cluster extraction unit clusters cells in which the radar received power exceeds the received power threshold and extracts the clusters. (1) A Doppler velocity variance calculation unit that calculates the Doppler velocity variance of each target cell within each cluster, or (2) a Doppler velocity variance deviation of the Doppler velocity variance of each target cell within each cluster with respect to the average Doppler velocity variance of surrounding cells outside each cluster. A cluster separation unit that (1) separates each target cell in each cluster from each cluster when the Doppler velocity variance of each target cell in each cluster is greater than or equal to the Doppler velocity variance threshold, and keeps each target cell in each cluster within each cluster when the Doppler velocity variance of each target cell in each cluster is less than or equal to the Doppler velocity variance threshold, or (2) separates each target cell in each cluster from each cluster when the Doppler velocity variance deviation of each target cell in each cluster is less than or equal to the Doppler velocity variance deviation threshold, and keeps each target cell in each cluster within each cluster within each cluster when the Doppler velocity variance deviation of each target cell in each cluster is greater than or equal to the Doppler velocity variance deviation threshold, This is a target image processing device characterized by having the following features.

[0009] With this configuration, considering that the Doppler velocity dispersion of the target is smaller than that of the system noise, it is possible to detect small, low-speed targets separately from the system noise, even when the small, low-speed targets are in close proximity to the system noise.

[0010] Furthermore, this disclosure further comprises a scan correlation calculation unit that calculates the scan correlation between past radar scans and current radar scans at each target cell within each cluster, The cluster isolation unit determines, based on the scan correlation of each target cell within each cluster, whether to separate or maintain each target cell within each cluster relative to the respective cluster. This is a target image processing device characterized by the following features.

[0011] With this configuration, considering that the scan correlation of low-speed targets is higher than that of clutter, even when low-speed, small targets are in close proximity to clutter, low-speed, small targets can be detected separately from the clutter.

[0012] Furthermore, the present disclosure states that the cluster separation unit determines whether to separate or maintain each target cell within each cluster from each cluster based on the scan correlation of each target cell within each cluster when the average Doppler velocity variance of the surrounding cells outside each cluster is below the average Doppler velocity variance threshold, and determines whether to separate or maintain each target cell within each cluster from each cluster based on the Doppler velocity variance of each target cell within each cluster when the average Doppler velocity variance of the surrounding cells outside each cluster is equal to or greater than the average Doppler velocity variance threshold. This is a target image processing device characterized by the following features.

[0013] With this configuration, when the average Doppler velocity variance of surrounding cells is small, like that of clutter, slow-moving, small targets can be detected separately from clutter based on the scan correlation of the target cell. On the other hand, when the average Doppler velocity variance of surrounding cells is large, like that of system noise, slow-moving, small targets can be detected separately from system noise based on the Doppler velocity variance of the target cell.

[0014] Furthermore, this disclosure is a target image processing program that causes a computer to execute each processing step performed by each processing unit of the target image processing apparatus described above.

[0015] This configuration makes it possible to provide a program that has the effects described above.

[0016] Furthermore, the inventions disclosed above can be combined as much as possible. [Effects of the Invention]

[0017] Thus, this disclosure enables the detection of small targets not only by using a radar device, but also by clustering cells where the radar received power exceeds a "small received power threshold" in order to detect small targets. This allows for the detection of small targets separately from system noise, even when the small targets are in close proximity to system noise.

Brief Description of the Drawings

[0018] [Figure 1] It is a diagram showing problems to be solved in the prior art's cluster extraction process. [Figure 2] It is a diagram showing the configuration of the object image processing system of the present disclosure. [Figure 3] It is a diagram showing the solution means of the cluster extraction process of the present disclosure. [Figure 4] It is a diagram showing the configuration of the object image processing system of the first embodiment. [Figure 5] It is a diagram showing the principle of the cluster separation process of the first embodiment. [Figure 6] It is a diagram showing the configuration of the velocity dispersion threshold processing unit of the first embodiment. [Figure 7] It is a diagram showing the results of the cluster extraction process of the prior art and the first embodiment. [Figure 8] It is a diagram showing the problems of the cluster separation process of the first embodiment. [Figure 9] It is a diagram showing the configuration of the object image processing system of the second embodiment. [Figure 10] It is a diagram showing the content of the scan correlation process of the second embodiment. [Figure 11] It is a diagram showing the content of the received power threshold processing of the second embodiment. [Figure 12] It is a diagram showing the configuration of the velocity dispersion threshold processing unit of the second embodiment. [Figure 13] It is a diagram showing the content of the Z-axis coordinate extraction process of the second embodiment. <​​​​​​​​​​​ (Configuration of the target image processing system in this disclosure) Figure 2 shows the configuration of the target image processing system of this disclosure. The target image processing system S comprises a radar device 1, a target image processing device 2, and a radar display device 3. The radar device 1 transmits a radar irradiation signal and receives a radar reflection signal. The target image processing device 2 comprises a target image acquisition unit 21, a cluster extraction unit 22, a radar characteristic calculation unit 23, and a cluster separation unit 24. The target image processing device 2 can be implemented by installing a target image processing program for performing the processing shown in Figures 6, 10 to 13 of the first and second embodiments on a computer. The radar display device 3 displays the position, velocity, and tracking results of the target.

[0021] Figure 3 shows the solution for the cluster extraction process in this disclosure. In this disclosure as well, it is necessary to cluster cells in which the radar received power exceeds a "small received power threshold" in order to detect not only large targets such as ships, but also small targets such as ships V or buoys. However, when small targets such as ships V or buoys are close to system noise N, it is not possible to detect the small targets such as ships V or buoys separately from the system noise N.

[0022] Here, the Doppler velocity dispersion of a small vessel V or a target such as a buoy is smaller than that of the system noise N. Therefore, in this disclosure, the Doppler velocity dispersion of each target cell (target or system noise N) within each cluster is calculated, and when the Doppler velocity dispersion of each target cell (system noise N) within each cluster is greater than or equal to the Doppler velocity dispersion threshold, each target cell (system noise N) within each cluster is separated from each cluster (target).

[0023] As a result, when tracking a small vessel V or buoy or other target, the target is not integrated with the system noise N, and the characteristic quantities of the tracked target (center of gravity position, illumination area, maximum received power, or average Doppler velocity, etc.) are not distorted, thus improving the accuracy of target tracking.

[0024] In Figure 3, small vessels V or targets such as buoys are detected separately from system noise N. However, as a modified example, targets such as small helicopters (whether their rotors can be detected) or aircraft may also be detected separately from system noise N.

[0025] (Configuration of the target image processing system in the first embodiment) Figure 4 shows the configuration of the target image processing system according to the first embodiment. The cluster extraction unit 22 includes a radar received power calculation unit 221 and a received power threshold processing unit 222. The radar characteristic calculation unit 23 includes a Doppler velocity dispersion calculation unit 231 and a velocity dispersion threshold processing unit 232. The cluster separation unit 24 includes a coordinate information extraction unit 241 and a clustering unit 242.

[0026] The principle of the cluster separation process in the first embodiment is shown in Figure 5. The left column of Figure 5 shows the XY plane coordinate information of the radar received power, and it is not possible to separate and detect the low-speed, small target when it is close to the system noise. The right column of Figure 5 shows the XY plane coordinate information of the Doppler velocity dispersion, and the Doppler velocity dispersion of the target is smaller than that of the system noise (the Doppler velocity of the system noise is near 0). Based on the right column of Figure 5, it is considered that even when a low-speed, small target is close to the system noise, it is possible to separate and detect the low-speed, small target.

[0027] Figure 6 shows the configuration of the velocity dispersion threshold processing unit of the first embodiment. The target image acquisition unit 21 acquires a target image from the radar device 1. The radar received power calculation unit 221 calculates the radar received power in the target image. The received power threshold processing unit 222 clusters cells in the target image where the radar received power exceeds the received power threshold using CFAR (Constant False Alarm Rate) processing (see Figure 12) and DBSCAN (Density-Based Spatial Clustering of Applications with Noise) processing, and extracts the clusters.

[0028] The Doppler velocity variance calculation unit 231 and the velocity variance threshold processing unit 232 use only the target cell TC in the CFAR processing, and do not use the surrounding cell PC and guard cell GC. The Doppler velocity variance calculation unit 231 calculates the Doppler velocity variance of each target cell TC within each cluster.

[0029] The velocity dispersion thresholding unit 232 outputs 0 as the Z-axis coordinate of each target cell TC in each cluster when the Doppler velocity dispersion of each target cell TC in each cluster is greater than or equal to the Doppler velocity dispersion threshold. On the other hand, the velocity dispersion thresholding unit 232 outputs 1 as the Z-axis coordinate of each target cell TC in each cluster when the Doppler velocity dispersion of each target cell TC in each cluster is less than the Doppler velocity dispersion threshold. The velocity dispersion thresholding unit 232 may set the Doppler velocity dispersion threshold to about half of the Nyquist velocity in the Fourier transform used to calculate the Doppler velocity (the Doppler velocity of system noise is near 0).

[0030] The coordinate information extraction unit 241 adds the Z-axis coordinate information of each target cell TC within each cluster to the XY plane coordinate information of the radar received power, thereby pseudo-expanding the coordinate information. The clustering unit 242 separates each target cell TC (system noise) within each cluster from each cluster (target) when the Z-axis coordinate of each target cell TC within each cluster is 0. On the other hand, the clustering unit 242 maintains each target cell TC (part of the target) within each cluster (target) when the Z-axis coordinate of each target cell TC within each cluster is 1.

[0031] Figure 7 shows the results of the cluster extraction process of the prior art and the first embodiment. The upper left column of Figure 7 shows the XY plane coordinate information of the radar received power, where a small, low-speed target is observed together with the system noise. The upper right column of Figure 7 shows the XY plane coordinate information of the Doppler velocity dispersion, where a small, low-speed target is observed separately from the system noise (the Doppler velocity of the system noise is near 0). The lower left column of Figure 7 shows the results of the cluster extraction process of the prior art (Patent Document 1), where the system noise is not separated from the small, low-speed target. The lower right column of Figure 7 shows the results of the cluster extraction process of the first embodiment, where the system noise is separated from the small, low-speed target.

[0032] Thus, considering that the Doppler velocity dispersion of the target is smaller than that of the system noise, it is possible to detect small, low-speed targets separately from the system noise, even when they are in close proximity to the system noise.

[0033] As a modification of the first embodiment, the Doppler velocity variance calculation unit 231 and the velocity variance threshold processing unit 232 may arrange the CFAR processing to reduce computational costs. The Doppler velocity variance calculation unit 231 may calculate the Doppler velocity variance deviation of each target cell TC within each cluster with respect to the average Doppler velocity variance of surrounding cells PC outside each cluster (outer cells of guard cells GC adjacent to the target cell TC) (calculated by the average Doppler velocity variance calculation unit).

[0034] Furthermore, the clustering unit 242 may separate each target cell TC (system noise) within each cluster from each cluster (target) when the Doppler velocity dispersion deviation of each target cell TC (system noise) within each cluster falls below the Doppler velocity dispersion deviation threshold. On the other hand, the clustering unit 242 may maintain each target cell TC (part of the target) within each cluster within each cluster (target) when the Doppler velocity dispersion deviation of each target cell TC (part of the target) within each cluster is equal to or greater than the Doppler velocity dispersion deviation threshold.

[0035] In the first embodiment, thresholding can be performed on the Doppler velocity variance of each target cell TC within each cluster, considering only the target region. On the other hand, in a modified version of the first embodiment, thresholding can be performed on the Doppler velocity variance deviation between each cluster and the average Doppler velocity variance of surrounding cells PC outside the cluster, considering the surrounding region as well.

[0036] Clutter can have a large Doppler velocity dispersion compared to targets, though not to the same extent as system noise. In the first embodiment, if the Doppler velocity resolution is sufficiently high, targets can be distinguished from clutter and system noise in thresholding that considers only the target region, and in the modified version of the first embodiment, if the surrounding region is also considered in thresholding that considers the surrounding region.

[0037] (Configuration of the target image processing system in the second embodiment) The problems with the cluster separation process in the first embodiment are shown in Figure 8. The left column of Figure 8 shows the XY plane coordinate information of the radar received power, and it is not possible to separate and detect the low-speed, small target when it is close to the clutter. The right column of Figure 8 shows the XY plane coordinate information of the Doppler velocity dispersion, and it is not possible to separate and detect the low-speed, small target when it is close to the clutter because the Doppler velocity of the clutter is not sufficiently random compared to the Doppler velocity of the system noise.

[0038] Figure 9 shows the configuration of the target image processing system according to the second embodiment. The cluster extraction unit 22 includes a radar received power calculation unit 221 and a received power threshold processing unit 222. The radar characteristic calculation unit 23 includes a Doppler velocity dispersion calculation unit 231, a velocity dispersion threshold processing unit 232, a radar received power storage unit 233 and a scan correlation calculation unit 234. The cluster separation unit 24 includes a coordinate information extraction unit 241, a clustering unit 242 and a Z-axis coordinate information extraction unit 243.

[0039] Figure 10 shows the scan correlation processing of the second embodiment. Figure 11 shows the received power threshold processing of the second embodiment. Figure 12 shows the configuration of the velocity dispersion threshold processing unit of the second embodiment. Figure 13 shows the Z-axis coordinate extraction processing of the second embodiment. In Embodiment 2, compared to Embodiment 1, the target image acquisition unit 21 and the radar received power calculation unit 221 perform similar processing, the received power threshold processing unit 222 performs slightly different processing, and the Doppler velocity dispersion calculation unit 231 and the velocity dispersion threshold processing unit 232 also perform slightly different processing.

[0040] In Embodiment 2, the radar received power storage unit 233 stores past radar scans after the output of the received power threshold calculation unit 222. The scan correlation calculation unit 234 calculates the scan correlation between past radar scans after the output of the received power threshold calculation unit 222 and the current radar scan after the output of the received power threshold calculation unit 222 for each target cell within each cluster.

[0041] As a variation, the radar received power storage unit 233 stores past radar scans prior to the input of the received power threshold calculation unit 222. The scan correlation calculation unit 234 calculates the scan correlation between the past radar scans prior to the input of the received power threshold calculation unit 222 and the current radar scan prior to the input of the received power threshold calculation unit 222 for each target cell within each cluster.

[0042] The scan correlation calculation unit 234 then determines that there is no temporal correlation when the radar reception intensity of each target cell within each cluster falls below a predetermined threshold in either a past radar scan or the current radar scan, and outputs 0 as the processing result (not the Z-axis coordinate) for each target cell within each cluster (see Figure 10).

[0043] On the other hand, the scan correlation calculation unit 234 determines that there is a temporal correlation when the radar reception intensity of each target cell within each cluster is above a predetermined threshold for both past and current radar scans, and outputs 1 as the processing result (not the Z-axis coordinate) for each target cell within each cluster (see Figure 10).

[0044] Here, system noise has almost no scan correlation, while clutter has some scan correlation, though not as much as target correlation. For example, clutter such as wind waves is irregular and therefore has little scan correlation, but clutter such as swells is regular and therefore has some scan correlation (see upper left column of Figure 11), and clutter such as rainfall has spread and therefore has some scan correlation (see lower left column of Figure 11).

[0045] Therefore, the received power threshold calculation unit 222 lowers the scan correlation of clutter by increasing the received power threshold, while maintaining a high scan correlation of targets. In other words, by increasing the received power threshold, the received power threshold calculation unit 222 fragments clutter such as swells and lowers the scan correlation (see the upper right column of Figure 11), and removes clutter such as rainfall overall and lowers the scan correlation (see the lower right column of Figure 11).

[0046] In Embodiment 2, the Doppler velocity variance calculation unit 231 and the velocity variance threshold processing unit 232 can reduce computational costs by arranging the CFAR processing. The Doppler velocity variance calculation unit 231 calculates the Doppler velocity variance of each target cell TC within each cluster and the average Doppler velocity variance of surrounding cells PC outside each cluster (outer cells of guard cells GC adjacent to the target cell TC) (calculated by the average Doppler velocity variance calculation unit AV).

[0047] The velocity dispersion thresholding unit 232 outputs 0 as the processing result (not the Z-axis coordinate) for each target cell TC in each cluster when the Doppler velocity dispersion of each target cell TC in each cluster falls below the Doppler velocity dispersion threshold. On the other hand, the velocity dispersion thresholding unit 232 outputs 1 as the processing result (not the Z-axis coordinate) for each target cell TC in each cluster when the Doppler velocity dispersion of each target cell TC in each cluster is equal to or greater than the Doppler velocity dispersion threshold. The velocity dispersion thresholding unit 232 may set the Doppler velocity dispersion threshold to approximately half of the Nyquist velocity used in the Fourier transform for calculating the Doppler velocity.

[0048] The velocity dispersion thresholding unit 232 outputs 0 as the processing result (not the Z-axis coordinate) for each peripheral cell PC outside the cluster when the average Doppler velocity dispersion of each peripheral cell PC outside the cluster falls below the average Doppler velocity dispersion threshold. On the other hand, the velocity dispersion thresholding unit 232 outputs 1 as the processing result (not the Z-axis coordinate) for each peripheral cell PC outside the cluster when the average Doppler velocity dispersion of each peripheral cell PC outside the cluster is equal to or greater than the average Doppler velocity dispersion threshold. The velocity dispersion thresholding unit 232 may set the average Doppler velocity dispersion threshold to about half of the Nyquist velocity used in the Fourier transform for calculating Doppler velocity, and it may be set to the same value or a different value compared to the Doppler velocity dispersion threshold.

[0049] The Z-axis coordinate information extraction unit 243 outputs the Z-axis coordinate of each target cell TC within each cluster based on the calculation result of the scan correlation of each target cell TC within each cluster when the threshold processing result of the average Doppler velocity variance of surrounding cells PC (clutter) outside each cluster is 0. In other words, the Z-axis coordinate information extraction unit 243 outputs 0 as the Z-axis coordinate of each target cell TC (clutter) within each cluster when the calculation result of the scan correlation of each target cell TC (clutter) within each cluster is 0. On the other hand, the Z-axis coordinate information extraction unit 243 outputs 1 as the Z-axis coordinate of each target cell TC (part of the target) within each cluster when the calculation result of the scan correlation of each target cell TC (part of the target) within each cluster is 1.

[0050] The Z-axis coordinate information extraction unit 243 outputs the Z-axis coordinate of each target cell TC within each cluster based on the threshold processing result of the Doppler velocity variance of each target cell TC within each cluster when the threshold processing result of the average Doppler velocity variance of the surrounding cells PC (system noise) outside each cluster is 1. In other words, the Z-axis coordinate information extraction unit 243 outputs 1 as the Z-axis coordinate of each target cell TC (part of the target) within each cluster when the threshold processing result of the Doppler velocity variance of each target cell TC (part of the target) within each cluster is 0 (it inverts 0 to 1 and outputs it). On the other hand, the Z-axis coordinate information extraction unit 243 outputs 0 as the Z-axis coordinate of each target cell TC (system noise) within each cluster when the threshold processing result of the Doppler velocity variance of each target cell TC (system noise) within each cluster is 1 (it inverts 1 to 0 and outputs it).

[0051] The coordinate information extraction unit 241 adds the Z-axis coordinate information of each target cell TC within each cluster to the XY plane coordinate information of the radar received power, thereby pseudo-expanding the coordinate information. The clustering unit 242 separates each target cell TC (clutter or system noise) within each cluster from each cluster (target) when the Z-axis coordinate of each target cell TC within each cluster is 0 (an indicator of low target likelihood). On the other hand, the clustering unit 242 maintains each target cell TC (part of the target) within each cluster (target) when the Z-axis coordinate of each target cell TC within each cluster is 1 (an indicator of high target likelihood).

[0052] Figure 14 shows the results of the cluster extraction process in the first and second embodiments. The upper left column of Figure 14 shows the XY plane coordinate information of the radar received power, where a small, low-speed target is observed together with the clutter. The upper middle column of Figure 14 shows the XY plane coordinate information of the Doppler velocity dispersion, where a small, low-speed target appears to be observed together with the clutter. The upper right column of Figure 14 shows the XY plane coordinate information of the scan correlation, where a small, low-speed target is observed separately from the clutter. The lower left column of Figure 14 shows the cluster extraction process result in the first embodiment, where the clutter is not separated from the small, low-speed target. The lower right column of Figure 14 shows the cluster extraction process result in the second embodiment, where the clutter is separated from the small, low-speed target.

[0053] Thus, considering that the scan correlation of low-velocity targets is higher than that of clutter, it is possible to detect small, low-velocity targets separately from clutter even when they are in close proximity.

[0054] Furthermore, when the average Doppler velocity variance of surrounding cells is small, like that of clutter, low-velocity, small targets can be detected separately from clutter based on the scan correlation of the target cell. On the other hand, when the average Doppler velocity variance of surrounding cells is large, like that of system noise, low-velocity, small targets can be detected separately from system noise based on the Doppler velocity variance of the target cell.

[0055] Furthermore, by considering that the scan correlation of clutter is somewhat high and using received power thresholding to keep it low, it is possible to detect small, low-speed targets separately from clutter even when they are in close proximity.

[0056] As a modification of the second embodiment, the Z-axis coordinate information extraction unit 243 may output the Z-axis coordinates of each target cell within each cluster, based not only on the threshold processing result of the Doppler velocity variance of each target cell within each cluster, but also on the calculation processing result of the scan correlation of each target cell within each cluster, when the threshold processing result of the average Doppler velocity variance of surrounding cells outside each cluster is 1.

[0057] In the second embodiment, when parts of low-speed targets overlap in past radar scans and current radar scans, the low-speed targets can be prevented from being excessively separated. Furthermore, in the second embodiment, when a high-speed target approaches a location where noise was present in a past radar scan during the current radar scan, the high-speed targets can be prevented from being excessively separated. On the other hand, in a modified version of the second embodiment, when a high-speed target approaches a location where a low-speed target was present in a past radar scan during the current radar scan, the high-speed targets can be separated from the low-speed targets. [Industrial applicability]

[0058] The target image processing apparatus and target image processing program of this disclosure, when clustering cells in which the radar received power exceeds a "small received power threshold" in order to detect not only large targets but also small targets using a radar apparatus, can detect small targets separately from system noise even when the small targets are in close proximity to system noise. [Explanation of symbols]

[0059] V: Ship N: System noise S: Target image processing system 1: Radar equipment 2: Target Image Processing Device 3: Radar display device 21: Target image acquisition unit 22: Cluster extraction unit 23: Radar characteristic calculation unit 24: Cluster Isolation Unit 221: Radar reception power calculation unit 222: Received power threshold processing unit 231: Doppler velocity dispersion calculation unit 232: Velocity dispersion threshold processing unit 233: Radar receiving power storage unit 234: Scan Correlation Calculation Unit 241: Coordinate information extraction unit 242: Clustering section 243:Z-axis coordinate information extraction part TC: Target cell GC: Guard Cell PC: Peripheral Cells AV: Average Doppler velocity dispersion calculation unit

Claims

1. A target image acquisition unit that acquires target images from a radar device, In the aforementioned target image, a cluster extraction unit clusters cells in which the radar received power exceeds the received power threshold and extracts the clusters. (1) A Doppler velocity variance calculation unit that calculates the Doppler velocity variance of each target cell within each cluster, or (2) a Doppler velocity variance deviation of the Doppler velocity variance of each target cell within each cluster with respect to the average Doppler velocity variance of surrounding cells outside each cluster. A cluster separation unit that (1) separates each target cell in each cluster from the cluster when the Doppler velocity variance of each target cell in each cluster is greater than or equal to the Doppler velocity variance threshold, and keeps each target cell in each cluster within the cluster when the Doppler velocity variance of each target cell in each cluster is less than or equal to the Doppler velocity variance threshold, or (2) separates each target cell in each cluster from the cluster when the Doppler velocity variance deviation of each target cell in each cluster is less than or equal to the Doppler velocity variance deviation threshold, and keeps each target cell in each cluster within the cluster when the Doppler velocity variance deviation of each target cell in each cluster is greater than or equal to the Doppler velocity variance deviation threshold, A target image processing apparatus characterized by comprising:

2. The system further comprises a scan correlation calculation unit that calculates the scan correlation between past radar scans and current radar scans for each target cell within each cluster, The cluster isolation unit determines, based on the scan correlation of each target cell within each cluster, whether to separate or maintain each target cell within each cluster relative to the respective cluster. The target image processing apparatus according to claim 1, characterized in that

3. The cluster separation unit determines whether to separate or maintain each target cell within each cluster from the respective cluster based on the scan correlation of each target cell within each cluster when the average Doppler velocity variance of surrounding cells outside each cluster falls below the average Doppler velocity variance threshold, and determines whether to separate or maintain each target cell within each cluster from the respective cluster based on the Doppler velocity variance of each target cell within each cluster when the average Doppler velocity variance of surrounding cells outside each cluster is equal to or greater than the average Doppler velocity variance threshold. The target image processing apparatus according to claim 2, characterized in that

4. A target image processing program for causing a computer to perform each processing step performed by each processing unit of the target image processing apparatus according to any one of claims 1 to 3.

Images