JP2002082169A - Radar image processing method and guidance system for guided body - Google Patents

Abstract

PROBLEM TO BE SOLVED: To shorten the pattern matching time of a radar image-processing method which converts the position of a specific body in an image of a ground radar into a target position in a guided body radar image through the pattern matching between images of the ground radar and a guided body radar. SOLUTION: When a body to a guided missile (2) is specified among plural bodies in the radar image obtained from the ground radar (1), target groups in the radar images of the ground radar (1) and guided body radar (3) is represented as contours, their contour coordinates are compared to find the position shift of the radar image, and a target position specified in the guided body radar image is detected. Only the coordinate points forming the contours need to be compared, and the processing time can be shortened greatly. Furthermore, the contours of the target groups are compared, so that a specific target hit rate nearly equal to that of a conventional template comparison system is obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a radar image processing method for pattern matching between a ground radar image and a derivative radar image, and a derivative guidance system.
The present invention relates to a radar image processing method for performing pattern matching at high speed and a derivative guidance system.

[0002]

2. Description of the Related Art A guidance system for guiding a derivative flying in the air from the ground is widely used for controlling flying objects such as missiles (guided bullets). For example, in an automatic guidance system that guides a guided missile to a target target, a target is specified for each guided missile from a plurality of targets detected by ground radar,
The guided bullet automatically tracks this designated target. At this time, the target position in the radar image obtained by the ground radar is different from the position viewed from the guided bullet radar equipped with a guided bullet at a position different from the ground radar. Therefore, by comparing the radar image obtained by the ground radar with the radar image obtained by the radar equipped with a guided bullet, the guided bullet detects the specified target.

At this time, since a position error occurs between the two radar images due to a positional shift or a difference in radar acquisition time, a specified target cannot be detected by simple collation. Meanwhile, 2
To detect the exact misalignment of two patterns,
A pattern matching method using a template is known. According to this method, similarity is inspected for all points on the image, paying attention to the area of the pattern and the point of interest, so that the pattern position shift can be detected accurately. As this method, a method of forming a reference image by a contour line is also known (for example,
JP-A-9-81747, etc.).

[0004]

However, since the similarity is checked for all points on the image, matching takes a long time when the number of elements constituting the image is large. That is, when a plurality of targets are used as elements like a radar image, processing time is required if template matching is performed at all points of the radar image using one of the radar images as a template. , The target position is designated, and the tracking ability is reduced. This is particularly noticeable when guiding a plurality of guided bullets.

An object of the present invention is to provide a radar image processing method and a derivative guidance system for speeding up pattern matching of a radar image without deteriorating the positional deviation detection accuracy.

It is another object of the present invention to provide a radar image processing method and a derivative guidance system for simplifying a radar image pattern matching algorithm and shortening the processing time.

[0007]

In order to achieve this object,
The radar image processing method of the present invention includes a step of extracting a contour line of an object group in a radar image of a ground radar and a derivative radar mounted on the derivative, and comparing the contour coordinates of the two contour lines, The method includes the steps of: obtaining a displacement amount of an object group in the radar image; and detecting a target position specified on the derivative radar image based on the displacement amount.

In addition, a guidance system according to the present invention comprises a derivative having a derivative radar, a ground radar for acquiring radar images of a plurality of objects, and a plurality of objects in a radar image obtained from the ground radar. A radar image processing device for specifying a target object to the derivative, wherein the radar image processing device extracts a contour line of a group of objects in a radar image of the terrestrial radar and a derivative radar mounted on the derivative, The outline coordinates of both outlines are collated to determine the amount of displacement of the object group of both radar images, and the target position specified on the derivative radar image is detected based on the amount of displacement.

According to the present invention, when specifying a target object from a plurality of objects in a radar image obtained from a terrestrial radar to a derivative, a target group in the radar images of the terrestrial radar and the derivative radar is represented by a contour line. By comparing the contour coordinates, the position deviation of the radar image is obtained and the object position specified on the derivative radar image is detected, so that only the coordinate points forming the contour need to be compared, and the processing time is greatly reduced. Can be shortened to In addition, since the contours of the object group are collated, the specified target hit ratio equivalent to that of the conventional template collation method can be obtained.

In the present invention, preferably, the collating step includes a step of detecting central coordinates of each of the two target groups from contour coordinates of the two target groups, and a provisional position based on a difference between the two central groups. Calculating the shift amount, and comparing the outline coordinates of the two outlines with the provisional position shift amount as an initial value, and determining the position shift amount of the object group of both radar images, The provisional displacement amount P1 is obtained from the difference (CA-CB) between the center coordinates of the object group, and the displacement amount between the contour point coordinates is calculated based on the provisional displacement amount P1. Can be calculated by

Further, in the present invention, preferably, the contour line extracting step includes a step of extracting a contour line of a target group in the radar image, and the target position detecting step includes a step of extracting a target specified by the derivative laser. Is detected.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described with reference to an embodiment of a guide system for a missile, but the present invention is not limited to this embodiment.

FIG. 1 is a block diagram of a guided bullet guidance system according to an embodiment of the present invention, FIG. 2 is an explanatory diagram of a pattern matching process, FIG. 3 is a flowchart of a target position detection process, and FIG.
3 is a flowchart of the collation processing in FIG. 3, and FIG. 5 is a flowchart of the contour detection processing in FIG.

As shown in FIG. 1, in an automatic guidance system for guiding an airborne guided missile 2 to a target target, a target is designated for each guided missile 2 from a plurality of targets 5 detected by the ground radar 1. The guided bullet 2 automatically tracks this designated target. At this time, the position of the target 5 in the radar image obtained by the ground radar 1 is different from the position of the guide bullet 2 located at a different position from the ground radar 1, that is, from the guide bullet radar 3 mounted on the guide bullet 2. Different from the position you saw. Since the guided missile 2 automatically tracks the position of the target detected by the guided missile radar 3, it is necessary to match the target position of the radar image obtained by the ground radar 1 with the target position of the radar image of the guided missile radar 3. There is.

For this reason, by comparing the radar image obtained by the terrestrial radar 1 with the radar image obtained by the guided bullet mounted radar 3, a matching process 6 for detecting the amount of positional deviation.
Is provided. A target point obtained by adding the amount of displacement to the specified target position of the target obtained from the radar image of the ground radar 1 is transmitted as collation information to the guidance bullet 3, and the guidance bullet 3 is obtained from the guidance bullet radar 3. Of the target positions, the target position closest to the target point is detected, and the
Finds the specified target.

The principle of the algorithm of the collation processing 6 is as follows.
This will be described with reference to FIG. FIG. 2A shows an example of a radar image, in which a black circle is a target. Ground radar 1 and guided missile radar 3
Means that the same target group 5 is acquired as a radar image, but the target position in the radar image is different because the radar position is different. In the collation processing 6, a contour line surrounding the target group detected by the ground radar 1 is acquired as coordinate data, and is compared with all points of the contour line coordinate data surrounding the target group in the radar image obtained from the radar 3 equipped with the guided bullet. Then, the similarity (correlation) is checked, and the position having the highest similarity is detected.

In the case of the target radar image example shown in FIG. 2A, as shown in FIG. 2B, a surface image of the target group is obtained from FIG. ), The contour of the target group surface image is detected. At this time, the detection of the contour line
According to the algorithm shown in (1), the coordinate value of each point is obtained.
At this time, the contour length is also obtained.

In the matching algorithm shown in FIG. 3 described later, first, the contour coordinates (G ₀ to G ₀ ) of the target group on the ground radar image are obtained.
G _n ). Similarly, the contour coordinates (Y _{0 to} Y _m ) of the target group on the guided bullet radar image are obtained. At this time, the contour length is also calculated. Second, the center coordinates of the target group are obtained from the maximum and minimum coordinates of the contour line. If there are multiple outline blocks,
A contour block having a contour length closest to the target group specified by the ground radar 1 is selected. Third, ground radar images, the difference between the center coordinates of the target group of missile radar image, the interim positional deviation amount P _1. Fourth, collation is performed for all points of the contour line. Thus, the amount of positional deviation is obtained.

The procedure will be described with reference to FIG.

[0020] (S1) determine the face image of the target group on the ground radar image, we obtain the outline coordinates of the target group on the ground radar images from the surface image (G ₀ ~G _n). Similarly, the plane image of the target group on the guided bullet radar image is obtained, and the outline coordinates (Y _{0 to} Y _m ) of the target group on the guided bullet radar image are obtained from the plane image. At this time, the contour length is also calculated. The contour detection algorithm is shown in FIG.
It will be described later. FIG. 2 (A) shows a surface image from the target group shown in FIG.
As shown in (B), in order to obtain the image, a so-called filling process is used in which the image is scanned in one direction, the area between the points is filled, the other direction is scanned, and the area between the points is similarly filled. .

(S2) In both the ground radar image and the guided bullet radar image, the center coordinates CA and CB of each target group are obtained from the maximum and minimum coordinates of the contour line. When there are a plurality of contour blocks, the contour block having the contour length closest to the target group specified by the ground radar 1 is selected.

(S3) The difference (CA-CB) between the center coordinates of the target group of the ground radar image and the target bullet radar image is set as a provisional displacement amount P1.

(S4) As described with reference to FIG. 4, all points of the contour are collated to determine the amount of positional deviation.

(S5) A point obtained by adding the amount of displacement to the designated target position of the ground radar image is the target point. The target closest to this target point is the specified target.

Next, the collation processing of FIG. 3 will be described with reference to FIG.

(S10) One point of (contour point coordinates + temporary positional deviation amount P1) obtained from the ground radar image is compared with all points Y0 to Ym of the contour point coordinates obtained from the guided bullet radar image. A point Y′0 at which the distance becomes minimum is obtained. For example,
G ₀ + P ₁ coordinates and Y ₀ to Y _m are compared, and the distance is the minimum.
Seek Y _'0.

(S11) A positional deviation amount P2 between one point of (contour point coordinates + temporary positional deviation amount P1) and a point Y'0 at which the distance between the contour point coordinates obtained from the guided bullet radar image is minimum is calculated. Comparing the (contour point coordinates + temporary position shift amount P1 + P2) obtained from the ground radar image with all the points Y0 to Ym of the contour point coordinates obtained from the guided bullet radar image, obtaining the number of points where the coordinates match, and recording I do. That is, (G ₀ + P ₁ + P ₂ ) to (G _n +
P ₁ + P ₂ ) and Y ₀ to Y _m . The number of points where the coordinates match is determined and recorded.

(S12) The number of detected matching points is compared with the number of past matching points, and if the number of detected matching points is large, the true displacement is updated to (P ₁ + P ₂ ). It is not updated unless the number of detected matching points is the maximum.

(S13) It is determined whether or not the displacement P2 has been determined for all the contour points determined from the ground radar image.
If so, end. If not, step S1 is performed to process the contour points obtained from the next ground radar image.
Return to 0.

In the above example of the contour points, first, G₀+ P₁Coordinate
And Y₀~ Y_mAnd a point Y 'at which the distance becomes the minimum₀Seeking
First, the position deviation amount P2 from the point Y'0 is calculated, and (G₀+ P₁+
P_Two) ~ (G _n+ P₁+ P_Two) And Y₀~ Y_mCompare. Coordinates match
Obtain and record the score to be given.

Next, similarly, update the G0 + P ₁ to G1 + P _1, it compares the G ₁ + P ₁ coordinate and Y ₀ to Y _m, determine the point Y _'1 whose distance is the smallest. G ₁ + P ₁ and Y ' ₁
And P _2, compares the _{G 0 + P 1 + P 2} ~G n + P 1 + P 2 and Y ₀ to Y _m, obtains a score coordinates match, compared to the past match scores, true the more consistent scores Set the displacement to P ₁ + P ₂ and G ₁ + P ₁
To update to the G ₂ + P _1. Do this until the G _n + P _1. Finally, the amount of positional deviation when the contour line most matches is obtained, and this is set as the true amount of positional deviation (P1 + P2).

As described above, when specifying a target from a plurality of targets in a radar image obtained from a ground radar to a guided missile,
The target group in the radar image of the ground radar and the guided radar is represented by a contour line, and by comparing the contour coordinates, the displacement of the radar image is obtained, and the target position specified on the guided bullet radar image is determined. To detect. Only the coordinate points forming the contour need be compared, and the processing time can be greatly reduced. In addition, since the contours of the target group are collated, a specified target hit ratio equivalent to that of the conventional template collation method can be obtained.

Further, in both the ground radar image and the guided bullet radar image, the center coordinates CA and CB of each target group are obtained from the maximum and minimum coordinates of the contour line, and the difference (C
A-CB) is used to determine the provisional displacement P1 and to calculate the displacement between the contour point coordinates on the basis of the provisional displacement P1, so that a highly accurate displacement can be calculated in a short time.

Next, referring to FIG. 5, a description will be given of the processing for detecting the contour line in FIG. This processing is well-known in, for example, "Basic of Image Processing" (published by Sangyo Hyoronsha Ltd. in 1986), and will be described briefly.

(S20) Scanning initial coordinates (x
(i, yj) is set to (x0, y0) = (0, 0).

(S21) The scanning is continued, one pixel without the tracked mark is found, a tracked mark is attached to this one pixel (boundary point) S0, and the tracking start point Sn = S0
Set to. If there is no one pixel without the tracked mark, the process ends.

(S22) As shown in FIG. 5 (B), for eight pixels around the starting point Sn as the center, when n = 0, the pixel with the number 1 starts, and when n> 0, the pixel with the number 1 starts. From the pixel of No. 6, the counterclockwise check is performed, and the first black pixel encountered is defined as Sn + 1.
The next boundary point. If there are no points (black pixels) in the eight pixels around Sn, the point is an isolated point and is a noise element.
It returns to step S21.

(S23) Sn + 1 = S1 and Sn = S0
It is determined whether the condition is satisfied. If this condition is satisfied, the process returns to step S21. On the other hand, when this condition is not satisfied, n is updated to n = n + 1, the process returns to step S22, and tracking is continued.

In this way, all points of the contour point coordinates can be obtained from the ground radar image and the guided bullet radar image. Further, the present invention can be used not only for guiding missiles and the like, but also for guiding other flying objects.

Although the present invention has been described with reference to the embodiment, various modifications are possible within the scope of the present invention.
They are not excluded from the scope of the present invention.

[0041]

As described above, according to the present invention,
When identifying a target object from multiple objects in a radar image obtained from a terrestrial radar to a derivative, the objects in the radar image of the terrestrial radar and the derivative radar are represented by outlines, and their outline coordinates are collated. Thus, the displacement of the radar image is obtained, and the target position specified on the derivative radar image is detected. Only the coordinate points forming the contour need be compared, and the processing time can be greatly reduced. In addition, since the contours of the object group are collated, the specified target hit ratio equivalent to that of the conventional template collation method can be obtained.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a guide bullet guiding system according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram of the collation processing in FIG. 1;

FIG. 3 is an overall flowchart of the collation processing in FIG. 1;

FIG. 4 is a flowchart of a collation process in FIG. 3;

FIG. 5 is a flowchart of the contour line extraction processing in FIG. 3;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Ground radar 2 Derivative 3 Derivative radar 4 Ground processor 5 Target 6 Collation processing

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G06T 7/60 300 G06T 7/60 300A // G01S 13/66 G01S 13/66 F term (Reference) 2C014 DA08 DB01 DC01 DD01 DD11 5J070 AC01 AF01 AF07 AK22 BD01 5L096 BA20 EA45 FA62 FA69 GA08 JA11

Claims (6)

[Claims] 1. A radar image processing method for specifying a target object from a plurality of objects in a radar image obtained from a terrestrial radar to a derivative, wherein the radar image of the terrestrial radar and a derivative radar mounted on the derivative is included. Extracting the contours of the object group of the object group; collating the contour coordinates of the two contour lines to obtain the amount of displacement of the object group of both radar images; Detecting the target position specified above. 2. The collating step includes: detecting a center coordinate of each of the two target groups from contour coordinates of the two target groups; and calculating a provisional positional deviation amount from a difference between the two center coordinates. 2. The radar according to claim 1, further comprising the step of: comparing the outline coordinates of the two outlines with the provisional deviation amount as an initial value to determine the amount of deviation of the object group of both radar images. Image processing method. 3. The contour line extracting step includes a step of extracting a contour line of a target group of the radar image, and the target position detecting step includes a step of detecting a target specified by the derivative laser. The radar image processing method according to claim 1, wherein: 4. A derivative having a derivative radar, a ground radar for acquiring radar images of a plurality of objects, and a target object is specified from the plurality of objects in the radar image obtained from the ground radar to the derivative. A radar image processing apparatus, wherein the radar image processing apparatus extracts contours of an object group in a radar image of the ground radar and the derivative radar mounted on the derivative, and acquires contour coordinates of the two contours. A derivative position of the object group of the two radar images, and a target position specified on the derivative radar image is detected based on the displacement amount. 5. The radar image processing device detects center coordinates of each of the two target groups from contour coordinates of the two target groups, and calculates a provisional positional shift amount from a difference between the two center coordinates. 5. The guidance system for a derivative according to claim 4, wherein the provisional positional deviation amount is set as an initial value, and the outline coordinates of the two outlines are collated to determine the amount of positional deviation of the object group of both radar images. 6. The derivative guidance system according to claim 5, wherein said radar image processing device extracts a contour line of a target group in said radar image and detects a target specified by said derivative laser.