JPH04196983A - Motion detector - Google Patents

Abstract

PURPOSE:To allow the compensation meeting a photographer's will at the time of motion correction by providing a filter for extracting the high-frequency range-component of a video signal before a circuit for detecting motion vectors from the video signal. CONSTITUTION:The output of a high-pass filter 8 which receives the output of a camera part 1 is suppressed in the component corresponding to a background of large motion and is made into the signal extracted with the component of an object which is the photographing object although the motion is small. Then, the motion is eventually detected from the video signal corresponding to the object which is the larger signal component in a motion vector detecting circuit 4 and, therefore, the motion vectors corresponding to the motion within the screen of the object is of course outputted. A part of a memory 3 is outputted by executing address control so as to shift the reading out address of the memory 3 in which the signals for one frame are previously stored by as much as the movement of the object. The video signal subjected to the motion correction is obtd. in this way.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to a motion detection device, and particularly to a motion detection device suitable for application to a motion compensation device for correcting shaking of a subject image in a television camera or the like.

[Prior Art] Conventionally, in the field of imaging technology using television cameras, a shake correction device for electronically correcting shake of a subject image within a screen due to camera shake has been known. In order to perform this shake correction, it is first necessary to detect the magnitude and direction of movement of the subject image within the imaging screen, and a motion detection device is used for this purpose. Figure 3 shows, for example, the document "About screen shake correction device" (Television Society Technical Report Vol. 11 No. 3).
4'3r-48 page, published in May 1989) is a block diagram of a conventional motion detection device applied to the shake correction device. In the figure, (1) focuses a subject image on an image sensor (1b) using a lens (1a), and
A camera unit that amplifies the image signal obtained from 1a) through a preamplifier (1c) and outputs it as a video signal, and (2) an A/D that converts the video signal derived from camera unit (1) into a digital signal. (analog/digital) transformer, (3) is a memory for writing and reading digital video signals from the A/D transformer), and (4) is the screen movement based on the video signal from the camera unit (1). a motion vector detection circuit that detects the direction and magnitude of the motion vector detection circuit (5), and a memory readout control that controls the readout address of the digital video signal from the memory (3) according to the motion signal detected by the motion vector detection circuit (4). circuit, (6) is memory (3
) is an interpolation circuit that enlarges and performs interpolation processing on a part of the motion-compensated digital video signal read out from the interpolation circuit (6). ) is a D/A (digital/analog) converter.

Next, the operation of the configuration described above will be explained.

The light image from the subject is captured by the lens (1) in the camera section (1).
A) forms an image on the imaging surface of the image sensor (1b), and the image sensor (1b) converts it into an electrical video signal. This signal is amplified by the preamplifier (1c) and the camera section (
1) is derived as an analog video signal. The video signal output from the camera section (1) is sent to the A/D converter (2).
) is converted into a digital signal at a constant sampling frequency and stored in the memory (3). Further, the motion vector detection circuit (4) detects the motion (velocity) in units of frames from the correlation between two temporally consecutive frames from the video signal. The memory read control circuit (5) reads the memory (3) based on the amount of motion determined by the motion vector detection circuit (4).
) is moved and converted, and a part of the image signal written in the memory (3) is read out. The partially read image signal is enlarged and interpolated by the interpolation circuit (6) to return to the normal size screen, and then converted to an analog video signal (video signal) by the D/A converter (7) and output. be done.

On the other hand, FIG. 4 is a block diagram showing the internal configuration of the motion vector detection circuit (4) in the configuration of FIG. 3. In the figure, (41) is an A/V converter that converts the analog video signal a output from the camera unit (1) into a digital video signal.
D (analog/digital) converter, (42), (4
5) is a latch circuit for determining the data of the digitally converted video signal, (43) is a representative point memory for storing and reading out the representative points of the pixels of the video signal, and (44) is a representative point memory (43). A latch circuit (46) calculates the absolute value of the difference between the representative point data b that has passed through the latch circuit (44) and the representative point data C that has passed through the latch circuit (45). absolute value circuit, (4
7) is a cumulative addition table that cumulatively adds the output data d of the absolute value circuit (46), and (48) is a cumulative addition table (48) that cumulatively adds the output data d of the absolute value circuit (46).
7) is a table value comparison circuit that compares the cumulatively added data.

The motion vector detection circuit (4) focuses on some pixels, that is, representative points, in a predetermined area on the screen in order to detect the movement of the image between frames of the video signal, and calculates the movement of the image from the amount of movement of these pixels. The method used is to determine the motion vector. The operation of this motion vector detection circuit (4) will be explained with reference to the explanatory diagram of FIG. By the way,
FIG. 5 shows the relationship between image blocks and representative points in the processing in the configuration of FIG. 4. In the figure, (51) is one field screen in the video signal, and (51) is one field screen in the video signal.
3) is a block obtained by dividing one field screen (51) into a predetermined number of blocks, (54) is a representative point Ri,j set at the center of each block (53), and (55) is a block obtained from each block. All pixels R1-1, which constitute each of (53),
jl~Ri+x, j+y.

Now, in the configuration shown in FIG. 4, the input video signal a is digitally converted by the A/D converter (41).
As a result, the block (
53) is written into the representative point memory (43) through the latch circuit (42). Representative point memory (4
The data stored in 3) is read out after being delayed by one frame, and is inputted to the absolute value circuit (46) as representative point data b via the latch circuit (44). On the other hand, the representative point Ri,j (54) of the video signal from the A/D converter (41)
) is input to the absolute value circuit (46) as representative point data C via the latch circuit (45).

That is, the representative point data b of the previous frame and the representative point data d of the current frame are input to the absolute value circuit (46), and the absolute value of the difference between them is calculated. These calculations are performed for each block (53) constituting one field screen (51), and the output data d of the absolute value circuit (46) has a table corresponding to the same address in each block (53). They are added one after another to the cumulative addition table (47). The addition result of this cumulative addition table (47) is input to the table value comparison circuit (48), and by finally knowing the minimum address of the value of the addition result, the direction and amount of movement of the image after one frame, i.e. The motion vector value is obtained as the output e of the table value comparison circuit (48).

In other words, the absolute value of the difference between the representative point Ri, j and the signal 5i+) (, j+y) located in the horizontal direction X1 and vertical direction y is determined, and the relationship between X and y in the same horizontal and vertical directions at each representative point is calculated. The absolute value of the difference at a certain point is added to form a cumulative addition table Dxy (47). At this time, Dxy is Dxy-Σl Rij-8j+x, j"y l
Indicated by i, j ・ ・ ・ (1). Then, the minimum value of X and y in this Dxy
Let be the horizontal and vertical motion vectors.

[Problems to be Solved by the Invention] Since the conventional motion detection device is configured as described above, there are a plurality of objects at different distances from the camera unit (1) within the same screen. ) swings in parallel within a plane whose normal is the direction in which the desired subject is located,
In addition to the subject, each object is detected as causing a different amount of screen shake. For this reason, there is a problem in that the unexpected screen shake of the object is corrected, and the screen shake of the intended subject cannot be completely suppressed.

Similarly, when there is no shaking in the camera unit (1) and part of the screen is moving, or when multiple objects are moving differently, unexpected objects may appear. There is a problem in that the movement is corrected. On the other hand, there are many problems that need to be solved, such as the problem that the subject of motion vector detection changes over time depending on the shooting situation, causing malfunctions such as parts of the screen moving or being fixed.

This invention was made to solve the above-mentioned problems, and when a camera section is taking an image of a target object, generally the target object is in focus, and other objects are not in focus. An object of the present invention is to provide a motion detection device that can reliably detect the motion of a target object by paying attention to the fact that the target object is off.

[Means for Solving the Problems] In order to achieve the above object, the present invention provides an imaging means for capturing an image of a subject through a lens system and transmitting a video signal consisting of a plurality of temporally continuous frames; a filter means for extracting a high-frequency component that is a focused portion, that is, a component corresponding to an object, from the video signal from the image pickup means; and an image containing many components corresponding to the object obtained by the filter means. The present invention provides a motion detection device including a detection means for calculating the movement of a screen from the movement amount and direction of representative points of pixels constituting at least two temporally continuous screens of signals.

[Operation] In the above means, the motion detection device of the present invention captures a subject image through a lens system in the imaging means to obtain a video signal consisting of a plurality of temporally continuous frames;
From this video signal, a focused portion, that is, a high frequency component corresponding to the target object, is extracted through a filter means, and a detection means is used to extract a temporally continuous video signal containing many components corresponding to the target object. The movement of the target object is detected by calculating the movement of the screen from the amount and direction of movement of representative points of pixels constituting at least two screens.

[Example] Hereinafter, the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram of a motion detection device according to an embodiment of the present invention. In the same figure, (8) is the lens (1a)
The image of the subject is formed on the image sensor (1b), and the image signal obtained from the lens (1a) is sent to the preamplifier (IC).
Camera unit (1) that amplifies and outputs as a video signal through
The output of this bypass filter (8) is input to the motion vector detection circuit (4). The other configurations are substantially the same as those shown in FIG. 3, and the configuration of the motion vector detection circuit (4) is also as shown in FIG. 4.

Next, the operation of the above-described configuration will be explained with reference to the explanatory diagram of the screen shown in FIG.

Now, the screen imaged by the camera unit (1) has a tree in the background (21)
Consider a case where a person standing as a person is imaged as an object (22). Furthermore, it is assumed that the photographer operates the camera unit (1) while walking, and the camera shakes while taking pictures. In this case, the focus of the camera unit (1) is naturally on the object (22), which is the subject, and the background (21) is in focus.
) becomes the out-of-focus point. Now, assuming that Fig. 2 (a) and (b) are two consecutive frames, the camera part (1) shakes downward in the screen of Fig. 2 (b) compared to the screen of Fig. 2 (a). It can be seen that it is something like this.

In this case, since the background (21) is far away and the target object (22) is nearby, the distant object, that is, the background (21), moves more than the target object (22) in the screen.

Now, the video signal obtained from the camera section (1) naturally has a component corresponding to the background (21) and a component corresponding to the object (22). The corresponding signal has more high-frequency components, and the signal corresponding to the out-of-focus background (21) has fewer high-frequency components. Therefore, in the output of the bypass filter (8) which receives the output of the camera unit (1), the component corresponding to the background (21) with large movement is suppressed, and the component corresponding to the background (21) with large movement is suppressed, and the component corresponding to the object (22) which is the object of photographing although there is little movement is suppressed. The component becomes the extracted signal. therefore,
Since the motion vector detection circuit (4) detects motion from the video signal corresponding to the object (22), which is a larger signal component, it is natural to detect the motion vector corresponding to the movement of the object (22) in the screen. Output. This motion vector is given to the memory readout control circuit (5), which receives the motion amount and receives the video signal from the camera unit (1) through the A/D converter (2). Address control is performed to shift the readout address of the memory (3), which stores the digitized signal for one frame, by the amount of movement of the object (22).
Output part of 3). Then, the interpolation circuit (6) applies interpolation processing to a part of the screen to enlarge it to the original size, and the D/A converter (7) converts it into an analog signal, resulting in a motion-corrected image. I can get a signal. As a result, the target object (
22) can be motion-corrected to the same position as in FIG. 2(a) to obtain a still screen.

In the above embodiment, motion detection between two frames has been described as an example, but the present invention is similarly applicable to motion detection between two or more frames. Further, in the above embodiment, a case is illustrated in which high-frequency components are extracted by passing an analog video signal through a bypass filter, but it is also possible to apply digital filter processing to a digital video signal to obtain a desired signal. .

[Effects of the Invention] As described above, according to the present invention, by simply providing a filter for extracting high-frequency components of a video signal in front of a circuit that detects a motion vector from a video signal, it is possible to capture a subject in focus, In other words, it is possible to extract and detect only the movement of the object, and even when performing motion correction, it is possible to perform compensation according to the photographer's intention, and a highly reliable system can be obtained.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a motion detection device according to an embodiment of the present invention, FIG. 2 is an explanatory diagram of a screen for explaining the operation of the configuration shown in FIG. 1, and FIG. 3 is a diagram of a conventional motion detection device. FIG. 4 is a block diagram showing the internal configuration of the motion vector detection circuit of FIG. 3, and FIG. 5 is an explanatory diagram showing the relationship between image blocks and representative points in processing in the configuration of FIG. 4. In the figure, (1) is a camera unit, (2) is an A/D converter, (3) is a memory, (4) is a motion vector detection circuit, (
5) is a memory read control circuit, (6) is an interpolation circuit, (
7) is a D/A converter, (8) is a bypass filter, (2
1) is the background, (22) is the object, (41) is the A/D converter, (42) is the latch circuit, (43) is the representative point memory,
(44) is a latch circuit, (45) is a latch circuit, (46)
) is the absolute value circuit, (47) is the cumulative addition table, (48
) is a table value comparison circuit, (51) is one field screen, (53) is a block, (54) is a representative point, and (55) is all pixels. In addition, in the figures, the same reference numerals indicate the same or equivalent parts.

Claims (1)

[Claims] an imaging means for imaging a subject image and transmitting a video signal consisting of a plurality of temporally continuous screens; a filter means for extracting a high frequency component from the video signal from the imaging means;
The present invention is characterized by comprising a detection means for calculating the movement of the screen from the movement amount and direction of pixels constituting at least two temporally continuous screens of the video signal containing many high-frequency components obtained by the filter means. Motion detection device.