JPS59135379A - Image multiple-object separating circuit - Google Patents

Abstract

PURPOSE:To enable the realization of a firearms control apparatus or a missile guiding apparatus capable of simultaneously tracking a plurality of targets and the application of the titled circuit to an apparatus for measuring or acquiring a plurality shapes in a real time, by performing that a plurality of targets in an image are mutually separated by numbering in a real time. CONSTITUTION:In a region number control part 12 in which a multi-value video signal is received, the video signal binary-coded by a threshold is stored as the content 8 of a region number control register and multiplied by target detection windows A-E to control region numbering according to the logic of a specific formula. Because it is necessary to apply a new number in a picture element satisfying a certain specific formula, an order is issued to a new region number generating part 13 from the region number control part 12 to generate a new region number which is, in turn, transmitted to and stored in a first register 14 as well as stored in a register line 15 through the first register 14 while sent to a region finish detecting part 19 and a flag for detecting region finish is erected.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an apparatus for tracking multiple targets by processing images obtained by imaging the targets, and more particularly to a circuit for separating and numbering multiple targets.

Generally, methods for tracking multiple targets using this type of image tracking device have been proposed, and a method for separating multiple targets is to search the boundary line between the target and the background in the circumferential direction of the target. Methods for separating targets have also been generally proposed. However, in the proposed target separation, 1 field (16.7 m5 for standard TV scan images)
) or more delay time is required.

Referring to the drawings below, FIG.
FIG. 1(a) is a diagram for explaining the outline of numbering targets in a dimensional processing area, and FIG. 1(a) is an example of a binarized image;
Figure (b) is an example of a numbered image, Figure 1 (c) is a target detection window, Figure 1 Fdl is an area number detection window, (1) is a processing area of a two-dimensional image obtained by raster scanning, (2) is the first binarization target assuming that two targets exist in the processing area (1), (3) is the second binarization target, and (4) is the second binarization target. ) after assigning number 1 to the binarization target, (5) assigns number 2 to the second binarization target.
is the second separation target after numbering.

Multiple targets (including clutter, etc.) existing within processing area (1) are numbered, and targets that are not connected are assigned different numbers to separate the multiple targets. , Here, for convenience of explanation, a method of numbering 2 will be explained for the case where there are two targets.

The data in the processing area (1) is a binary video signal that is input sequentially from the upper left to the lower right in a raster scan manner.

The first binarization target (
2) and the second binarization target (3) are numbered and identified as separate groups. When data is input sequentially from the upper left using sister scanning, first the first binarization target (
2) appears.

A negative sign 1 is attached to this mass to make it the first separation target (4), and since a 2-negative target (2) appears next, the number 2 is added to it and it is made the second separation target (5).

From the first window to the fifth window in Figure 1 (C) (indicated by A to E in the figure, hereinafter referred to as A to B), the window 2 in Figure 1 (a) is
The numbering of the darkened pixels in the first window is controlled over the digitized image.

Then, from the first window to the fifth window (a in the figure) in Figure 1(d)
From a to e, hereinafter represented by a to e - Go. ) as the first
The number of the pixel covered by the first window is selected from among the second window to the fifth window and set in the numbered image shown in FIG.

For example, if only A is 1 and B, O, D, and E are 0, then the number of pixel a in the window in Figure 1(d) is
Get a new number that is not in use. If the pixel of A is l and there is a pixel that takes 1 out of B, ("3.D, H, then there is a pixel connected to A, and it has already been numbered, so The first one corresponding to that pixel
The numbers stored in windows s, c, d, and e in Figure (C) are conveniently used as the number of the pixel in a. The window in Figure 1(C) is applied to the binarized image in Figure 1(a), and the window in Figure 1(d) is applied simultaneously to the numbered image in Figure 1(d). It will be done.

In this way, the two targets are numbered as the first separation target (4) and the second separation target (5), but in reality even within one target there are irregularities. Different numbering will be performed for many targets, and a detailed method for solving these problems will be explained using FIG. 2.

Figure 2 is a diagram for explaining the method of assigning one number to one target, and Figure 2 (a) is a diagram for explaining a state in which a target detection window is applied to a binarized image. , Figure 2(b)
(6) is a number 1 start point where area number 1 is named first, and (7) is a diagram to explain the state in which area number detection windows are applied to numbered images. Number 2 start point to be named, (8) is the content of the area number control register that stores information for one horizontal quantitative line before the pixel indicated by the white letter A in the target detection window, (9)
is the content of the area number line register that stores information for one horizontal scanning line before the pixel indicated by the white letter a in the area number detection window, and (10) is the content of the area number line register that stores information for one horizontal scanning line before the pixel indicated by the white letter a in the area number detection window. U is the number rewriting point where it becomes necessary to replace one number with another after realizing that the number has been changed, and U is the area end detection point where it is detected that the numbering of one target has been completed.

The point represented by A in the target detection window is the current pixel, B is the pixel one pixel past, and 0. D and E are pixels one horizontal line past the pixel of A, and the pixels from A to E move sequentially from the upper left to the lower right of the screen with their relative positions fixed. 21N(a) Part 1
It shows how it moves from left to right during the horizon period. Furthermore, the area detection window also moves in the same way as the target detection window. The target detection window and the area detection window simultaneously move corresponding pixels on the screen, and the writing and rewriting of the incrementing area number is controlled by looking at the value of the binarized image and determining the value of the target detection window. It is controlled by the values of the five windows A-E.

The value of the window a is determined based on the values of the five windows b to e of the area number detection window.

The logic of numbering is expressed by the following equations (1) to (5).

F(A)=C1→f(a)−〇・・・・・・・−・
...+11F(5)△F'(B)△F(0)ΔF's)△F[exist])=1→ f(a)=b...
・・・−・・・・・・(21F(A)ΔF(E)△
(Fσ3)V F(0) ) −1−+ f(a)−f
(b) Vf(c) -===-(3)F(a)ΔF's)ΔF(0)Δ(F's)F(E)) = 1-+
f(a)=f(d)Vf(e) −・−(4)F(A
)ΔF’s)ΔF(g))Δ(F(B)F(0)) =
1→f(a)=f(b)vf(c) -=・----
(5) When formula (5) holds true, f(x)=f(e) −f(x)−f(a) −=
−−−−−−−=−(6) Here, F(A) is the content of the area number control register in target detection window A (8)
and takes the value O or 1. (The same applies to F(g)) to F(g)) The entry represents the area number and is a positive integer.

f (a) is the content (9) of the area number line register in the area detection window a, and takes values from O to n, and Ω=
0 represents no area.

Also, f(X) is the content of the area number line register (9)
It is.

In the target detection window and area detection window (hereinafter referred to as window) where there is no target on the screen, the area number is written as O, as expressed by equation (1). When the window reaches the starting point (6) with number 1, the area number is 1, as expressed by equation (2).
Add. Next, at the point where the window advances by one pixel, only the values of F(A) and F() are 1 and F(C) to F(g)) are 0, so as shown in equation (3), f The value 1 of (b) is set to f
Assign to (a). When the window comes to the starting point (7) of number 2, only F(A) is 1 and F(B) to F(E) are O, so we assign the number to f(a) using equation (2). 1 Since number 1 is already in use, number 2 is newly adopted. In order to control the area number, the contents of the area number control register (8) as shown in the shaded area in Fig. 2(a).
, 1 in the region where the target exists and O elsewhere.
Remember. Similarly, regarding the area numbered image, the area number of each pixel on one horizontal line is stored as the content (9) of the area number line register, as shown by the hatched area in FIG. 2(b). When the window reaches the number rewriting required point αQ, area number 1 and area number 2 coexist within windows a to e.

In this case, F(4)=F(g))-F's)=1. F[
Since f(b)=2 is added to f(a) according to equation (5), and f(e
), which is equal to 1 in f (X) (content (9) of the area number line register), is replaced with the value 2 of f (a). After this, the number 1 is stored as an unused number in the vacant number register, and when assigning a new number, the contents of the vacant number register are checked, and if a vacant number exists, it is used with priority over the new number.

It is determined that the area has ended at the area end detection point υ, but in this method, a 7 lag is set for the area number used in the previous line, and then the area number used in the current line is set. Judgment is made by comparing flags.

In the case of Figure 2(b), the area number used in the previous line is only 2, and the area number used in the current line is O.
Therefore, the numbering of the area numbered with area number 2 has been completed.

Next, one embodiment of the present invention will be described.

In FIG. 3, α2 receives and receives a multilevel video signal;
Area number control unit that controls area numbering depending on the presence or absence of a target using a binary image after digitization, α3 is window A
A new area number generator, which is used when there is a target in the area but there are no other targets around it and a new numbering is required, (141
is the first register that stores the region number of the pixel in window a, (151 is a register string that stores the region number of one horizontal scanning line past window a, and +1G is for determining replacement of the region number. A fifth register, ση, for use as
Ua is a rewriting necessity detection unit that determines whether or not it is necessary to replace with one type of number when two or more types of area numbers are assigned to one target. When the need arises, the area number rewriting detection section searches for the same area number as the area number that needs to be rewritten in the contents of the register column (151). This is a region end detection section.

Area number control unit 0 that receives and receives multi-head video signals
In 21, the video signal binarized by the threshold is stored as the content (8) of the area number control register, and is multiplied by the target detection window A-g to obtain the values (1) to 21.
Control area numbering according to the logic of equation (5). It is necessary to assign new numbers to pixels that satisfy formula (2).

From the area number control section a'a to the new area number generation section (
13 to generate a new area number, transfer it to and store it in the first register a4, and store it in the register row a5 via the first register ■.
Area end detection unit u! Send it to J and set 7 lags to detect the end of the area.

In the area number control section αZ, the formula (3) or (
4) If there are pixels that have already been numbered around window a (window b-e) as expressed by the formula.

A switching signal for selecting a window (window be) for numbering the pixels of window a is sent to the area number a register C141.

The first register α-hiro stores the contents of windows b-e in register row a9.
, select one necessary piece of data, and store it as the contents of window a.

Rewriting necessity detection unit (Lη receives the contents of window e from register row port 9 via the fifth register (LSI), and compares the contents of window a from the first register with the area number 1. Determine the need for replacement.

The area number: replacement detection unit ae receives the finger and data indicating whether rewriting is necessary from the rewriting safety detection unit, looks at all pixels stored in the register array, and replaces the contents of pixels that require rewriting. The area number rewriting detection unit UIC reads the contents of the area number stored in the register 9 at a high speed corresponding to one pixel,
Search for the area number that is equal to the area number of window e and replace it with the area number of window a. The time required for replacement is the time required for one line to correspond to one pixel on the screen.

The data of the area number that has become unused due to the area number replacement is transferred to the rewriting necessity detection unit (from L71 to the new area number generation unit 0).
floor and make that area number reusable. at the same time,
The data of the unused area number is sent to the area end detection unit (I9) and the area number flag is changed.The area end detection unit α9 detects the area used in the current line when it reaches the end time of the horizontal line, that is, the right edge of the screen. The flag of the area number used in the previous line is compared with the flag of the area number used in the previous line to detect whether there is an unused area number. It is determined at the area end detection point αυ that the attachment has been completed, and this information is sent to the outside.

The numbering of regions for region division has been explained above, but when incorporating this into a multi-target tracking device etc.
If two numbers are replaced at the same time as the data is sent to the calculation unit (maximum video level, coordinates of the maximum video level (double centroid coordinates, vertical and horizontal length, etc.), the feature parameter calculation unit also sends the data for each area number. Control to replace.

Then, when the end of the region is detected, the characteristic parameter of the target to which the end number has been assigned becomes the final 1@, which is used as information for selecting a tracking target and tracking table.

Next, the numerical values of an example of actually implementing this circuit into hardware will be specified. Input image - The cycle for numbering the screen is 16.7ms, and the numbering delay time is 63.5μs
The processing screen is divided into 127 vertically and 31 horizontally, and the maximum number of detectable targets is -945 per screen (however, 2
The numbering is 128 due to the limitation of the subsequent circuit), and l
The maximum number of targets on a horizontal line is 15.

As described above, by using the multi-target separation circuit of the present invention, it becomes possible to realize a fire control system or missile guidance system that can simultaneously track multiple targets, and also to realize multiple shapes in real time. It can be applied to devices for measuring and understanding, and its effects are significant.

[Brief explanation of the drawing]

Figure 1 is a diagram to explain the outline of numbering targets in a two-dimensional processing area, Figure 2 is a diagram to explain how to assign one number to one target, Figure 3 is a diagram for explaining an embodiment of the present invention, (1) is a processing area, (2) is a first 2Ll@ target, (3) is a second bivalent target, and (4) is a diagram for explaining an embodiment of the present invention. is the first separation target, (5) is the second separation target, (6) is the start point of number 1, (7) is the start point of number 2, (8) is the content of the area number control register, (9 ) is the content of the area number line register, α
Q is the point where number rewriting is required, αυ is the area end detection point, n21
is the area number control section, 031 is the new area number generation section, α-yu is the first register, n51 is the register string, n
61 is a fifth register, αD is a rewriting necessity detection section, 0 seconds is an area number rewriting detection section, and σ9 is an area end calibration section. In the drawings, the same or corresponding parts are designated by the same reference numerals. Agent Shin Kuzuno - Figure 1 (Shi) (C) (f) Figure 2 (α) C Shin

Claims (1)

[Claims] Binarizes a Fuster scan video signal obtained from an image pickup device, video playback device, etc. and stores it in a register to control the area number, and assigns a number using the first to fifth windows for extracting the target. an area number control unit for controlling the area number control unit; a first register for selecting and storing the content of the area number to be written in the first window according to a command from the area number control unit; and the first register and the fifth register. a rewrite necessity detection unit that compares the contents of the area numbers from the area numbers and determines whether or not the area numbers need to be rewritten; a register column that stores the area numbers from the first register for one horizontal scanning line; An area number rewriting detecting unit that receives a command from the rewriting necessity detecting unit and replaces all pixels that are equal to the value of the fifth window in the contents of the register string, and detects that target numbering is completed. Consists of area end detection section 2
An image multi-target separation circuit characterized in that a plurality of targets within a screen are separated from each other in real time by numbering them.