JPS63187889A - Television telephone image pickup device - Google Patents

Abstract

PURPOSE:To realize a portrait image pickup device for television telephone where undesired part other than a person is not displayed by extracting a portrait image from a video signal from a television camera and forming an image onto the other part generated from a pattern generator prepared separately. CONSTITUTION:An object area identification circuit 100 identifies an area of an object having movement and an area of a standstill background and forms an object area identification signal (b) bringing the object area to a significant level '1' and the background area into an ineffective level '0'. A switch 510 is controlled to be thrown to the position of an input video signal (a) when the object area identification signal (b) is at level '1' and to be a position of a switch 500 when '0'. The video signal outputted from the switch 510 is a background synthesized video signal where the background is a curtain and an object image is inserted therein.

Description

DETAILED DESCRIPTION OF THE INVENTION (1) Further Fields of the Invention The present invention relates to a device for photographing a person's image in a videophone.

(2) Conventional technology The imaging device currently widely used uses an optical system 1 to convert an object 2 into a two-dimensional planar photoelectric conversion element 3 as shown in FIG.
The image of the subject formed on this photoelectric conversion element is
The video signal is extracted by electronically scanning t. Therefore, the optical system is adjusted so that the target subject is focused on the photoelectric conversion element. When such an imaging device is used for a videophone call or the like, although the image of one person is focused, the image formed on the photoelectric conversion element 3 is output as is, so the background etc. are transmitted as is to the other party. For this reason, there were major restrictions on installation, such as having to devise a location to place the device so that inconvenient objects would not be shown.

On the other hand, there is a technology that extracts moving subjects from video signals. Since the present invention utilizes this technology, Part 1
Let me give you an example first.

FIG. 4 shows an example of such a video extracting device, in which 5 is a switch, 10 is a background memory, 15 is a background difference calculation circuit, and 2
0 is a binarization circuit, 25 is a sampling signal production circuit, 30 is a video gate circuit, 35 is a background pixel extraction circuit, 41, 42
．． 43 is a delay circuit.

In this operation, first, the background pixel extraction circuit 35 calculates the difference between frames of the input video signal for each pixel, and extracts pixels with no difference. Extract the corresponding pixel. The switch 5 is based on the information extracted by the background pixel extraction circuit 35.

For the background pixel part in the input video signal, the input video signal side is accepted, and for the other pixel parts, the background memory 1 is accepted.
The contact point is controlled to accept the video signal side read from 0. Due to this.

Among the pixels of the video signal read from the background memory 10, the background pixel portion included in the human-powered video signal is replaced with pixel information of the input video signal.

input into the background memory 10; Through this operation, the latest background information is always stored in the background memory 10.

The background image stored in the background memory 10 is read out in synchronization with the input video signal, and is input to the background difference calculation circuit 15. The background difference calculation circuit 15 calculates the difference for each pixel between the background image inputted from the background memory 10 and the manually inputted image, and inputs the difference value of each pixel to the binarization circuit 20. 2
The value conversion circuit 20 generates a signal by converting the difference value inputted from the background difference calculation circuit 15 into a binary value using a certain threshold value. By the above operation, a binary signal having a significance level ("1") is obtained only in the pixel portion where a difference occurs between the input video signal and the background video signal read out from the background memory 10. This binarized signal has an invalid level "0" in a pixel portion inside the subject whose level is similar to that of the background. The sampling signal generation circuit 25 generates a video extraction signal that corrects the pixel portions inside the subject that have an invalid level of “0” to a significant level of “1,” and supplies the signal to the video gate circuit 30. The video gate circuit 30 passes only the portion of the input video signal in which the video sampling signal has a significant level of "1", and outputs a video signal with the other portions set to black level. Through the above operations, the video signal of only the desired subject image is extracted. The delay circuits 41, 42, and 43 correct the processing time in each circuit, and the delay circuit 41 corrects the processing time in the background pixel extraction circuit 35.

The purpose is to match the phase of the background pixel information output from the background pixel extraction circuit 35 and the input video signal, and the delay circuit 42 is used to match the phase of the background pixel information output from the background pixel extraction circuit 35 and the input video signal.
The delay circuit 43 is used to match the phase of the video sampling signal outputted from the sampling signal production circuit 25 and the input video signal delayed by the delay circuit 42. It is something.

In the present invention, the delay circuit 43 of the video extraction device
The portion from which the video gate circuit 30 is removed is referred to as a "subject area identification circuit", and the video sampling signal obtained from the sampling signal generation circuit 25 is used as the subject area identification signal.

(3) Object of the Invention An object of the present invention is to provide a person image capturing device for a videophone in which unnecessary parts other than a person, which is a target subject image, are not displayed.

(4) Structure of the Invention (4-1) Features of the Invention and Differences from the Prior Art The present invention takes advantage of the fact that a person is moving and captures images from a television camera obtained by scanning a photoelectric conversion element. The main feature is that the human image is extracted from the video signal, and the other parts are generated by a separately prepared screen generator.The video signal extracted by simply scanning a photoelectric conversion element is This is fundamentally different from conventional imaging devices that output data as is.

(4-2) Embodiment FIG. 1 is a diagram showing an embodiment of the present invention, in which 100 is a subject area identification circuit, 200 is an identification situation monitoring circuit, and 300 is a diagram showing an embodiment of the present invention.
．． 310 are the 1st. second background generation circuit, 400
5'00°510.520 is a video switching device, and 600 is a delay circuit. Also, Figure 2 shows the first
FIG. 1 is a diagram illustrating an example of the embodiment shown in FIG.
is the input video signal, b is the subject area identification signal, C or C'
is a background signal 5, d is a background composite video signal, e or e' is a foreground signal, and f, r' or f'' is a foreground embedded signal (output video signal).

A video signal a captured by the conventional television camera shown in FIG. 3 is input to a subject area identification circuit 100. The object area identification circuit 100 identifies a moving object area and a stationary background area.

For example, when referring to the subject area shown in Figure 2, the bell "l"
2. Create a subject area identification signal with the background area set to invalid level "0." A switch 510 switches the subject area identification signal to the human input video signal a side during a significant period (for example, level "1"), and switches it to the human video signal a side during a period when it is invalid ("0").
Control the contact to the 00 side. Now, assuming that the contacts of the switch 500 are always on the background generation circuit 1 side of the switch 300,
The output signal of the switch 510 is such that the subject portion remains the input video, and the other portions become the video generated by the background generation circuit 1 of 300.

For example, if the background generated by the background generation circuit 1 of 300 is a curtain image as shown in FIG. 2C.

The video signal output from the switch 510 becomes a background composite video signal in which the background is a curtain and the subject image is embedded in it, as shown in FIG. 5d. now.

If a desk or the like is not placed in front of the subject, this signal can be used as is as an output video signal.

As is clear from the above explanation, when the subject is not captured, the subject area identification signal sets the entire area as a motionless background area (level 10').

The video is only the video output from the background generation circuit 1 of 300, and the input video signal is not output at all. In addition, when a subject, for example a person, enters the input video signal, only the part of the subject becomes the input video signal, and the other part becomes the image of the background generation circuit 1 of 300. In either case, the input video signal The static background inside is not included in the output video signal. Note that the delay circuit 600 is for correcting the processing time in the subject area identification circuit 100, matching the phase of the input video signal a in the switching device 510 and the subject area identification signal S, and operating the switching device 510 correctly. be. Further, the background generation circuit 1 of 300 can be created by using video memory, computer graphics, etc., or most simply, a plain background such as blue can be used to create a desired background.

However, as shown in the example of FIG.
If a desk, etc. is shown in front of the subject in the image, the background composite video signal will be an image with the lower part of the subject cut off, as shown in Figure d, resulting in an unnatural image where the subject (person) is floating. Become. The foreground generation circuit 400 and the switch 520 are parts for eliminating this unnatural image.As shown in FIG. An identification signal is generated and supplied to the switch 520. Based on the foreground area identification signal from the foreground generation circuit 400, the cutter 520 controls the contact on the foreground generation circuit 400 side during the period of the foreground image portion, and on the switch 510 side during other periods. Through this operation, the video signal output from the switch 520 becomes a video signal (output video signal) in which the foreground is embedded in the background composite video signal d, as shown in FIG. 2r, and the unnatural feeling described above is removed. . This foreground generation circuit 400, like the background generation circuit 1 of 300, is created using video memory or computer graphics.
This can be realized by generating a foreground area identification signal as shown in 1).

In this case, it is easier to use computer graphics. Furthermore, by setting everything below a certain scanning line as the background, the contact point of the switch 520 is moved to the switch 510 side at the scanning start point (top left of the screen) of each frame of the video signal, and the background generation circuit 400 It is also possible to simplify the foreground region identification signal by switching to the side only once every frame. In addition, the foreground generation circuit 4
00 is a plain color (for example, brown) and the foreground is accented only in the horizontal direction, thereby making the circuit simple. Examples of the foreground signal e and the output video signal f in this simplified case are shown in Fig. 2 e' and f', respectively.

Next, the identification status monitoring circuit 200 monitors the object area identification signal outputted from the object area identification circuit 100, and detects the presence (“1”) indicating the object area in the object area identification signal over the entire five screens. ), the no-object signal is determined to be significant and is supplied to the switch 500. The switch 500 controls the contact to the background generation circuit 2 side of 310 when this no-object signal is significant, and to the background generation circuit 1 side of 300 during other periods. Through this operation, when the subject (person image) is not included in the input video signal, the background becomes the video generated by the background generation circuit 2 of 310. Therefore, if the content of the background generation circuit 2 of 310 is set as a message, the message can be sent when the subject is not shown. 310 in Figure 2 C'
An example of a screen generated by the background generation circuit 2 is shown in FIG.

Further, the no-object signal output from the identification status monitoring circuit 200 is also supplied to the switch 520.

By stopping the operation while the no-object signal is significant "1" and always switching to the switch 510 side, it is possible to make the entire screen the image of the background generation circuit 2 of 310. In this case, Messages can be displayed on the entire screen. It is also possible to prepare a plurality of messages so that they can be selected according to the situation, or to input any message using a personal computer or the like. Furthermore, each of the background generation circuits 1 and 2 and a plurality of foreground generation circuits may be provided, and by switching between different backgrounds and foregrounds and supplying them to the switching devices 500 and 520, the background and foreground can be selected. Not only is it possible, but it is also possible to make considerable changes simply by changing the color tone of the background and foreground. 2. 300 background generation circuits 1, 310 background generation circuits 2 and 4 foreground generation circuits used here
It goes without saying that .degree.O operates to output a video signal synchronized with the input video signal delayed by the delay circuit 6.degree.O input to the switch 510.

(5) Detailed Description of the Invention According to the present invention, a subject (such as a human image) is extracted from a video signal obtained from a conventional imaging device,
Only the part of the subject can be used as the input video signal, and the other part can be used as a separately prepared background or foreground screen, so you don't have to pay attention to the background and the background that you don't want to be seen will not be shown. By using the imaging device of the invention, there is an advantage that it is possible to provide a very convenient video telephone that does not require selection of a placement location.

[Brief explanation of the drawing]

Figure 1 is a diagram showing an embodiment of the device of the present invention, Figure 2 is a diagram explaining the operation of Figure 1, Figure 3 is a diagram explaining a conventional imaging device, and Figure 4 is a diagram used in the present invention. FIG. 2 is a diagram illustrating a conventional technique. 10o: Subject area identification circuit 200: Identification status monitoring circuit 300: Background generation circuit 1 310: Background generation circuit 2 400: Foreground generation circuit

Claims (2)

[Claims] (1) A means for scanning a subject image formed on a two-dimensional planar photoelectric conversion element and converting it into a video signal, a means for identifying a region of the subject image in the video signal, and a first and second means for generating a background, and monitoring the identification result of the area of the subject image, identifying that the subject is not shown, and when the subject is not shown, controlling the predetermined background generation means of the first and second background generating means; means for selecting one of the background generating means when the subject is present, and selecting the other background generating means when the subject is shown; and the area of the identified subject image is a video signal obtained by scanning the photoelectric conversion element; The other areas include means for converting the video signal generated by the selected background generation means into a background composite video signal, means for generating a foreground, and a means for converting the foreground generated by the foreground generation means into the background composite video signal. 1. A videophone imaging device comprising: embedding means. (2) In the scope of claim (1), one or two or all of the first background generation means, the second background generation means, and the foreground generation means each have a plurality of image generation means, and A video telephone imaging device characterized in that one of a plurality of video generating means is selected as first, second background and foreground video signals, respectively.