JPH0719286B2 - Motion detection device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a moving body detection device for extracting the movement of an object in color image processing, for example.

[Outline of Invention]

The present invention relates to a moving object detection apparatus, processing input data composed of a plurality of data series, extracting fixed pattern data for each series, comparing the extracted data and the input data for each series, and comparing these. By synthesizing the outputs and using this synthesized output to select the input data and a fixed value to obtain the moving object data, the moving object can be detected in the original image and the background is made a uniform monochromatic color, which is a good moving object. It enables detection.

[Conventional technology]

For example, when detecting the motion of a moving body in image processing, conventionally, for example, the difference between two time-series data is detected.

However, with this method, only the contour of the moving body can be obtained, and the contour in the traveling direction of the moving body is thick and the contour in the direction perpendicular to it is thin, and it is difficult to supplement the accurate image of the moving body. It was

By the way, the applicant of the present application has previously proposed a digital signal processing device (see Japanese Patent Laid-Open No. 58-215813) applicable to video image processing.

That is, FIG. 3 illustrates the outline of the device. In the figure, (21) is an input terminal and (22) is an input / output control (IO
C) system, (23) input image memory (VIM) system, (24) signal processing (PIP) system, (25) address generation (PVP) system,
(26) is the output image memory (VIM) system, (27) is the main control (T
C) system, (28) is an output terminal.

In this device, an analog video signal from a video camera (not shown) or the like is supplied to the input terminal (21). This video signal is supplied to the IOC system (22) and converted into predetermined digital data by AD conversion or the like, and the VIM system (2
Written in 3). In addition to digital data, signals for controlling the VIM system (23) are supplied from the outside of the IOC system (22) such as clocks, control mode signals, addresses, and write control signals.

The VIM system (23) is supplied with signals for controlling the VIM system (23) from inside the PVP system (25) such as the address of the digital data to be processed, write control, read mode, and data select. The address data is mutually transferred to the PIP system (24) for processing. Further, the data processed by the PIP system (24) is supplied to the VIM system (26), and the address etc. from the PVP system (25) are supplied to the VIM system (26).
The digital processed by this is written in the VIM system (26).

Further, the VIM system (26) is also supplied with an address or the like from the IOC system (22), the digital data read by this is supplied to the IOC system (22), and converted into a predetermined analog video signal by AD conversion or the like. It is converted and taken out to the output terminal (28).

The TC system (27) supplies a designation signal such as a mode and system, a clock signal, etc. to each system (22) to (26).

Further, a start signal of a frame to be processed is supplied from the IOC system (22) to the PVP system (25), and a processing end signal is supplied from the PVP system (25) to the IOC system (22).

In this way, the video signal supplied to the input terminal (21) is digitally processed and taken out to the output terminal (28). According to the above-described device, the functions required for processing are provided in each system. (22) ~ (26), each system (22) ~
Since a control circuit can be independently provided for each (26) and control can be performed by an independent microprogram, the load of software for each system is small and high-speed processing can be performed with a simple program. This makes it possible, for example, to process video signals in real time.

By the way, in the above device, the contents of the processing are
4) etc. determined by the microprogram. Therefore, the content of the processing can be changed by rewriting these microprograms.

That is, FIG. 4 shows a specific configuration of the PIP system (24),
This PIP system (24) is actually formed by arranging a large number (for example, 60) of processor units (30) in parallel,
In the figure, only two of them (30a) and (30b) are shown. In this figure, the digital data from the VIM system (23) or (26) is the processor unit (30a) (30b) ...
・ Input registers (FRA) (31a) (31 provided for each
b) ... and these registers are PVP
The system (25) controls the read addresses of the VIM systems (23) and (26) according to the read addresses, and stores a predetermined amount of data required for each processor unit.

The data written in these registers (31a) (31b) ... are the arithmetic units (32a) (33a), (32b) (33b), respectively.
Is supplied to ... Each of these arithmetic units is provided with an adder / subtractor, a multiplier, a coefficient memory, and a data memory, and performs linear and non-linear data conversion arithmetic operations according to control signals from the control units (34a) (34b). Further, this calculation result is obtained by the calculation units (33a) (33b) ..., and these calculation units (33a) (33b) ... are written in the VIM systems (23) (26) by the PVP system (25). It is controlled according to the address, and the operation result is written in the desired portion of the VIM system (23) (26).

In this case, the control signals from the control units (24a) (34b) ... Are microprogram memories (MPM) (35a).
(35b) is formed according to the microprogram written in. Therefore, the MPMs (35a) (35b) ... Have a so-called RAM configuration, and microprograms from the outside are written into the MPMs (35a) (35b) ... through the change units (36a) (36b). As a result, the contents of the processing can be changed by rewriting the microprogram.

The inventor of the present application pays attention to this point.

[Problems to be solved by the invention]

The above-described conventional technique has a problem that only the contour of the moving body can be obtained, and an accurate image of the moving body cannot be supplemented.

[Means for solving problems]

The present invention relates to a moving object detecting device having first, second and third processing circuits for respectively extracting moving object data from R, G and B signals which are three primary color signal components of an input image signal. Input signal (FRA (31R) (31G) (31
B)) is provided with a logical sum means (9) for outputting a control signal corresponding to the logical sum, and the first, second and third processing circuits output coefficient A and coefficient B respectively. (2R) (2G) (2B) and a predetermined input signal of the three primary color signal components is multiplied by the coefficient A and output.
Multiplication means (1AR) (1AG) (1AB), the output signal of the first multiplication means and the second multiplication means (1BR) (IBG) (1B)
Addition means (4) for adding the output signal of B) and storage means (3R) (3G) (3B) for storing the output signal of the addition means
And comparing means (5R) (5G) (5B) for detecting a level difference between the input signal and the output signal of the storage means and supplying a signal corresponding to the level difference to the logical sum means, and the input signal And data of a predetermined level (terminals (8R) (8G) (8B)) are supplied, and the level difference between the input signal and the output signal of the storage means is a predetermined value (terminal (6R) according to the control signal).
(6G) and (6B)), the selecting means outputs the input signal, and outputs the data of the predetermined level when the level difference between the input signal and the output signal of the storage means is a predetermined value or less. (7R), (7G), and (7B), and the second multiplication means multiplies the output signal of the storage means by the coefficient B and supplies the result to the addition means. The moving object detecting device is characterized in that the relationship between the value a of the coefficient A and the value b of the coefficient B stored in is selected so that a + b = 1.

[Action]

According to this, a fixed pattern in the input data is extracted for each data series, the fixed pattern is compared with the input data, the comparison output is synthesized, and the synthesized output is used to generate the input data and the constant value. Since the selection is made, the image of the moving object can be detected as it is, and the background can be made a uniform monochromatic color, so that good moving object detection can be performed.

〔Example〕

By the way, the inventor of the present application has previously proposed the following moving body detection device. In FIG. 2, the input data Xn from the FRA (31) is supplied to the multiplier (1A), and the coefficient memory (2)
Is multiplied by the weighting coefficient A from the data memory (3), and the accumulated data Y _n-1 from the data memory (3) is supplied to the multiplier (1B), which is also multiplied by the weighting coefficient B from the coefficient memory (2). Further, the signals from the multipliers (1A) and (1B) are supplied to the adder (4), and the added data is written in the data memory (3). The above weighting factors A and B are A + B = 1.

As a result, the addition data Yn newly written in the data memory (3) becomes Yn = AXn + BY _n-1 , and this data Yn is further fielded back to the multiplier (1B) as cumulative data, and this is repeated. The displacement pattern in the input data Xn is removed with
Only fixed patterns are extracted.

That is, in the above, the input data X _n and the accumulated data Y _n-1
The portion that is invariant (fixed pattern) between and is A + B = 1, so that the value is used as it is as the next accumulated data Yn. Change to this (fixed pattern-displacement pattern)
The portion that has been attenuated is attenuated at the ratio of the weighting factors A and B to be the next accumulated data Yn, and this displacement pattern is further attenuated according to the time ratio of the pattern to the accumulated time. As a result, the fixed pattern in the input data Xn is extracted. This fixed pattern corresponds to the background data Y in the input data Xn.

Further, the extracted fixed pattern data and the input data Xn from the FRA (31) described above are supplied to the comparator (5), and the difference between them is supplied to the terminal (6).
The comparison signal is taken out when it becomes larger. This comparison signal is supplied to the selector (7), the input data from the FRA (31) is selected only during the period when this comparison signal is being output, and at the other periods, the predetermined data from the terminal (8) is selected. The data of the value d is selected. This selected data is supplied to the VIM system (23) (26) as output data.

Therefore, in the method used in this device, the input data X is selected when the difference between the input data X and the fixed background data Y is equal to or larger than the threshold value ε, and the data having the predetermined value d is selected when the difference is equal to or smaller than the threshold value ε. Is From the selector (7), the data Z of the pattern of the moving body moving in the arbitrary background determined by the predetermined value d.
Is taken out.

Thus, according to the above-described device, the moving body can be detected as an object as well as its contour, so that an image of the moving body can be accurately obtained, and particularly since the background is uniformized to a predetermined value, the moving body can be detected well. You can

Here, the above-mentioned device is a case where the input data is one system. On the other hand, for example, when processing a color video signal, it is necessary to process data of a plurality of systems such as three primary colors (R, G, B). In that case, for example, the above-mentioned device 3
When the system is installed in parallel and each color is processed independently, if, for example, an arbitrary color component in a moving object is close to the background, only that color is not detected and the color of that part is reproduced correctly. There was a risk that it would not be done.

The present application has been made in consideration of such points.

That is, in FIG. 1, the configurations from the FRA (31) to the comparator (5) and the terminal (6) described above are provided in parallel for each system. In addition, suffixes (R, G, B) are attached in the drawing.

The comparison outputs of these comparators (5) are supplied to the logical sum circuit (9).

On the other hand, the input data from FRA (31R) (31G) (31B) is supplied to the selectors (7R) (7G) (7B) and the terminal (8R)
Predetermined values d _R , d _G and d _B from (8G) and (8B) are supplied to the selectors (7R) (7G) (7B). The OR output from the OR circuit (9) is the selector (7R) (7G) (7B).
FRA (31R) (31
Input data from G) (31B) is selected, and a predetermined value from the terminals (8R) (8G) (8B) is selected during the other periods. This selected data is the output data
It is supplied to the VIM system (23) (26).

Therefore, in this device, if even one comparative output of the data of three systems is output, the three primary colors are switched at the same time, and if any color in the moving body is close to the background, it is switched by other data. It is possible to eliminate the possibility that the color of the moving body in that part is not correctly reproduced because the above process is not performed.

In this way, the moving object is detected, but according to the above-mentioned device, even in the data processing of a plurality of systems such as a color image, there is no variation in the data processing, and the moving object can be detected extremely well. .

The algorithm of the above detection method is expressed by the following equation.

Although the moving object is detected in this manner, the above-described method allows the moving object to be detected very satisfactorily even in the data processing of a plurality of systems such as a color image without causing variations in the data processing. .

In the above example, the extraction of the fixed pattern (background) is performed by weighting the input data and the cumulative data and adding them.
This is for example As described above, a constant may be added or subtracted according to the magnitude of the input data and the accumulated data, and another fixed pattern extraction method may be adopted.

〔The invention's effect〕

According to the present invention, a fixed pattern in the input data is extracted for each data series, the fixed pattern is compared with the input data, the comparison output is synthesized, and the synthesized output is used to generate the input data and the constant value. Since the image of the moving object is detected as it is, the background is made into a uniform monochromatic color, and good moving object detection can be performed.

[Brief description of drawings]

FIG. 1 is a block diagram of an example of the present invention, FIG. 2 is a block diagram of another example, and FIGS. 3 and 4 are diagrams for explaining a conventional technique. (1AR) (1AG) (1AB) (1BR) (1BG) (1BB) is a multiplier, (2R) (2G) (2B) is a coefficient memory, (3R) (3G)
(3B) is a data memory, (4R) (4G) (4B) is an adder,
(5R) (5G) (5B) are comparators, (7R) (7G) (7B) are selectors, and (9) is an OR circuit.

Claims (1)

[Claims] 1. A moving body detecting apparatus having first, second and third processing circuits for extracting moving body data from R, G and B signals which are three primary color signal components of an input image signal, respectively. A logical sum means for outputting a control signal corresponding to the logical sum of two input signals is provided, and the first, second and third processing circuits are respectively provided with a coefficient A.
And a coefficient storing means for outputting a coefficient B, a first multiplying means for multiplying a predetermined input signal of the three primary color signal components by the coefficient A and outputting the same, and an output signal of the first multiplying means. Adding means for adding the output signal of the second multiplying means,
Storage means for storing an output signal of the adding means; comparing means for detecting a level difference between the input signal and an output signal of the storage means; and a signal according to the level difference for supplying to the OR means. The input signal and the data of a predetermined level are supplied, and the input signal is output when the level difference between the input signal and the output signal of the storage means is larger than a predetermined value in response to the control signal. And a selection means for outputting the data of the predetermined level when the level difference between the output signals of the storage means is a predetermined value or less, and the second multiplication means has the output signal of the storage means and the coefficient. B is multiplied and supplied to the addition means, and the relationship between the value a of the coefficient A and the value b of the coefficient B stored in the coefficient storage means is selected so that a + b = 1. Is characterized by That the moving object detection apparatus.