JPS5933234B2 - Digital image tracking device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a digital image tracking device that tracks the center of gravity of a target using a digital image.

Usually, an image captured using an imaging device is A/D converted,
In a digital image tracking device that automatically tracks a target by calculating the center of gravity of an image of the target in the digital image (hereinafter abbreviated as target image), real-time performance is most important.

In order to process image data at high speed to meet this requirement, a large computer and a large capacity memory are essential. By the way, if this type of digital image tracking device is to be installed as a missile guidance device, there are restrictions in terms of size, weight, and power supply capacity, and naturally a large computer cannot be installed. Therefore, in the past, a small bipolar microcomputer was developed as a dedicated computer and processed in real time, but due to processing time constraints, it could only process images of about 100 pixels x 100 pixels at most. Furthermore, even though it was a dedicated microcomputer, it had the disadvantage of being about 300 x 200 x 100 mm in size and about 5 kg in weight, making it impossible to mount it on a small missile. In order to eliminate these drawbacks, the present invention constructs a weighting factor multiplication calculation unit in the X-axis and Y-axis directions of the target image, an area calculation unit, and a barycenter coordinate calculation unit in parallel digitally.
It consists of a comparator, a parallel adder circuit, and a counter circuit, and aims to speed up image data processing time, make the device smaller and lighter, and reduce power consumption.

A detailed explanation will be given below with reference to the drawings. Here, the case of 1-bit processing, that is, binarization, will be explained as an example of A/D conversion. FIG. 1 schematically shows a binarized image for explaining the barycenter coordinate calculation operation in this invention.

In FIG. 1, the barycentric coordinates XO, YO of the target image 2 in the binarized image 1 are given by equation (1), where the output of the pixel with coordinates i is expressed as Pij,,l''or,0''. It will be done. No. (
In formula 1), the numerator is the output of the weighting coefficient multiplication unit in the X-axis and Y-axis directions of the target image 2, and the denominator is the output of the area calculation unit.

Figure 2 shows the weighting coefficient multiplication operation in equation (1) and
FIG. 2 is a configuration diagram of an embodiment of a circuit for realizing Jrij, area calculation', and barycenter coordinate calculation.

In FIG. 2, the X-axis and Y-axis coordinate signal generation circuit 3a
and 3b to generate coordinate signals in synchronization with the scanning time of the binarized image 1 shown in FIG.

Each coordinate signal is sent to the gate circuit 5a according to the binarized image signal 4.
and 5b, and is input to parallel adder circuits 6a and 6b, and weighting coefficient multiplication calculation unit outputs 2'→'IPij and in the X-axis direction and Y-axis direction are obtained. On the other hand, by inputting the binarized image signal 4 to the counter circuit T and counting the number of pixels corresponding to the size of the target image, the area calculation unit output of the target image →→“Pij” is obtained.

After the above calculations of ゛ and →ヰPij are completed, the output of the area calculation section of l → '2' PIJ is input to the parallel adder circuit 8, and the clock signal from the clock generation circuit 9 is input by the Jl input LjJn calculation command →" →“The parallel digital comparator 10a generates Jl values that are integral multiples of Pij in sequence.
and 10b.

The parallel digital comparators 10a and 10b
Using , compare the integer multiples of the above and the integer multiples of RiJ in parallel, and check whether they are equal or not for each X-axis and Y-axis. →゛→”ゞWhen the integer multiple of Ij becomes large, a stop pulse signal is generated, and the addition instruction clock signal is sent to the counter circuits 11a and 11b.
When counting is stopped at , the counted values up to that point give the center of gravity coordinates Xc, Yc of the target image 2.

FIG. 3 is a time chart for explaining the above barycentric coordinate calculation operation.

In FIG. 3, the addition command clock signal (FIG. 3a) is generated at the same time as the weighting factor multiplication operation and the area operation are completed. When this addition command clock signal (Figure 3a) is counted until the generation of the stop pulse signals (Figure 3b) and (Figure 3c) in the X- and Y-axis directions, the count signal Nx (Figure 3d) and N, (Fig. 3e) is obtained. The values of Nx, n, obtained in this way are equal to the center of gravity coordinates Xc, Yc when the coordinate origin is set as shown in FIG. The weighting factor multiplication calculation unit, area calculation unit, and center of gravity coordinate calculation unit described above are collectively referred to as the center of gravity tracking calculation unit.The computation time of this center of gravity tracking calculation unit is limited by the comparison of parallel digital comparators. If a bipolar digital IC is used, the processing time Tc per comparison is several + Ns, and since the parallel bit comparison method is adopted as described above, the total calculation time is at most NxXTcns. Or NyXTcns 0 As an example, Texas
When using SN74LS85 manufactured by Instrument Co., Ltd., Tc is 36 ns, and the maximum value of Nx and N, is 2
When set to 56, the maximum processing time is TcXnx (or Ny) = 36ns x 256 = 9216ns, and this value is 33ms for one frame of a TV scanning type imaging device.
This corresponds to approximately 0.03% of

On the other hand, when trying to realize the above processing using a conventional microcomputer, it takes about 33ms, but by using the barycentric coordinate calculation section of the present invention, processing speed is about 3580 times faster than that of the conventional one. can.
Furthermore, the center of gravity tracking calculation unit described above can be configured based on a normal bipolar digital IC, and the number of parts required is small, making it compact, lightweight, and low power consumption. IC
can be easily converted into Although the above description has been made regarding the case where the device is installed as a missile guidance device, the present invention is not limited to this, and can be applied not only to tracking devices installed in aircraft, ships, vehicles, etc. and used outdoors, but also to semiconductor chips such as transistors and ICs. It can also be applied to measurement devices such as detection devices for positioning of automatic ponding devices.

As described above, in the digital image tracking device according to the present invention, the center of gravity of the target is calculated using the computer ττ−1.
! ” Divide using ττ −Jlj data ↓÷,
．． One and bitter←1. – instead of parallel digital
Using a dedicated processor consisting of comparators, etc. ΣΣX-P... ．． 11jJ1, integral multiple of gigi Pij, γgi YjPij and ΣΣP... ．． The coordinates of the center of gravity can be calculated simply by digitally comparing and calculating the integral multiples of 1J, which speeds up the calculation of the center of gravity and facilitates real-time processing of image data.It is also compact, lightweight, and has low power consumption. It has the advantage that it can be easily mounted on small missiles.

Further, since the circuit configuration is simple and the number of parts is small, it has the advantage that it can be easily converted into a hybrid IC.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a binarized image for explaining the barycenter coordinate calculation operation in this invention, and FIG. 2 is a diagram showing a circuit for realizing the weighting coefficient multiplication calculation section, the area calculation section, and the barycenter coordinate calculation section. FIG. 3, which is a block diagram of the embodiment, is a time chart for explaining the barycentric coordinate calculation operation. In the figure, 1 is a binarized image, 2 is a target image, 3a and 3b are X
Axis and Y-axis coordinate signal generation circuit, 4 is a binary image signal,
5a and 5b are gate circuits, 6a and 6b are parallel adder circuits, 7 is a counter circuit, 8 is a parallel adder circuit, 9 is an addition instruction clock generation circuit, 10a and 10b are parallel digital comparators, and 11a and 11b are counter circuits. be.

Claims (1)

[Claims] 1. Separate and extract the desired target image from the background by utilizing the difference in brightness of the image within the image obtained by capturing the desired target image and its background within the field of view of the imaging device, and extract the desired target image from the background. X-axis - Y set in the horizontal and vertical directions of
The imaging device calculates the barycentric coordinates of the target image in an axis-orthogonal coordinate system, detects an error signal between the barycentric coordinates and a predetermined coordinate, and controls the attitude of the imaging device so that this error signal becomes zero. In a digital image tracking device that tracks a desired target image so that it remains stationary at a predetermined position within the field of view,
A coordinate signal generation circuit that causes the target image barycentric coordinate calculation unit to generate coordinate signals X_i and Y_j in the X-axis and Y-axis directions for dividing the image obtained by the imaging device into pixels, which are the minimum units thereof; X-axis based on coordinate signal
Scan sequentially in the Y-axis direction and set the brightness of each pixel to "1" or "
Image P_i_j binarized to “0” (in case of “1”, P_i
If _j=1, “0”, convert A to P_i_j=0)
/D conversion circuit, add the above coordinate signals for each X-axis and Y-axis direction only when the above binarized image P_i_j is "1" and calculate the coordinate weighting calculation value ▲ There are mathematical formulas, chemical formulas, tables, etc. ▼
There are P_i_j and ▲mathematical formulas, chemical formulas, tables, etc.▼P_
Parallel addition circuit for calculating i_j, the above binarized image P_i_
The area calculation value of the pixel where j is "1" ▲ There are mathematical formulas, chemical formulas, tables, etc. ▼ Counter circuit to calculate P_i_j, addition instruction clock signal generation circuit to generate the clock signal to calculate the center of gravity coordinates, the above clock signal Based on the above area calculation value ▲ There are mathematical formulas, chemical formulas, tables, etc. ▼ P
A value that is an integer multiple of _i_j ▲ There are mathematical formulas, chemical formulas, tables, etc. ▼ Parallel addition circuit that calculates P_i_j (n = 1, 2, 3...), a value that is an integer multiple of this area calculation value ▲ Mathematical formula, chemical formula , tables, etc. ▼ Weighted calculation value of P_i_j and the above coordinates ▲ Numerical formulas, chemical formulas, tables, etc. ▼ P_i_j
and ▲ There are mathematical formulas, chemical formulas, tables, etc. ▼ P_i_j and ▲
There are mathematical formulas, chemical formulas, tables, etc. ▼Digital comparator that compares the magnitude relationship with P_i_j and generates a stop pulse signal in each X-axis direction and Y-axis direction when the following formula is satisfied, ▲Mathematical formulas, chemical formulas, tables, etc. There is ▼(1)▲
There are mathematical formulas, chemical formulas, tables, etc. ▼ (2) The above stop/
Until the pulse signal is generated, the numbers n_x and n_y of the clock pulse signals generated from the addition instruction clock signal generation circuit are counted in each of the X-axis direction and the Y-axis direction, and the barycenter coordinates (X_c, Y_c) of the following formula are calculated. ▲There are mathematical formulas, chemical formulas, tables, etc.▼(3)▲There are mathematical formulas, chemical formulas, tables, etc.▼
(4) A digital image tracking device characterized by comprising: a counter circuit for determining (4).