JPH099234A - Video conference system - Google Patents

Abstract

PURPOSE: To segment, magnify and display the image of a speaker from the video images of a television camera for projecting all conference participants by projecting the image of the speaker beforehand during zoom-up at the time of adjusting the direction of a camera. CONSTITUTION: Input images to a main television camera 5 arranged so as to project the conference participants are projected by an entire display television monitor 8 as they are. Also, the input images obtained by the camera 5 are sent to a moving body discrimination means 10 and the positions of the respective conference participants are discriminated. Further, a sub television camera 6 is arranged parallelly to the camera 5, the image of the camera 5 and the image of the camera 6 are compared and a distance from the television camera to an object is calculated in a distance calculation means 11. It is determined beforehand that the conference participant performs a motion such as raising a hand or the like before speaking. The speaker is discriminated in the means 10, position information decided in a speaker discrimination means 12 is sent to a controller 14, the camera 7 is focused, the speaker is acquired and the screen of the camera 7 is switched to from the image magnifying means 13 of the monitor 9.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is for smoothly switching screens of a video conference device, and a video camera for showing all participants in a conference until the zoom-up TV camera catches a speaker. The present invention relates to a video conferencing device that cuts out an image of a speaker from the video and displays the enlarged image.

[0002]

2. Description of the Related Art In a conventional video conference apparatus, when a video of only a speaker is displayed on a screen, a method of cutting out the image of the speaker from the images showing all the participants in the conference and a camera for zooming up are provided. There is a method of framing the speaker by moving. For the former, show all participants in the conference on a high-definition TV such as HDTV.
A method has been considered in which a speaker is judged based on a sound picked up from a microphone, and an image of the speaker is cut out and displayed on a conventional NTSC system television or the like.

Regarding the latter, a method in which the operator of the camera points the camera at the speaker by remote control, a method in which the television camera is automatically pointed at the speaker by a sound picked up from the speaker's microphone, and the like are considered.

[0004]

However, in such a video conference apparatus, in the method of cutting out the image of the speaker from the whole image, HDTV and standard definition TV are used.
It was necessary to prepare TV monitors and TV cameras with different resolutions. Further, in the method of moving the zoom camera, the image of the speaker cannot be immediately displayed while the camera is zooming and the direction is adjusted, and it is necessary to wait until the camera captures the speaker.

[0005]

SUMMARY OF THE INVENTION The present invention is to solve these problems, and in a video conference apparatus having a display means for displaying an object photographed by a camera, an operation for discriminating the operation of the object. Discriminating means, extracting means for extracting the subject discriminated by the motion discriminating means, image enlarging means for enlarging the image of the subject extracted by the extracting means, and photographing the image enlarged by the image enlarging means. Provided is a video conference device including a photographing means and a display means for displaying an image photographed by the photographing means.

[0006]

According to the present invention, it is possible to smoothly switch the screen of a video conference by displaying the image of the speaker in advance while zooming up while adjusting the orientation of the camera.

[0007]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a basic configuration of a video conference apparatus of the present invention, FIG. 2 is a block diagram of the video conference apparatus of the present invention, and FIG. 3 describes a distance calculating means of the video conference apparatus of the present invention. FIG. 4 is a diagram for explaining the image enlarging means of the video conference apparatus of the present invention. In FIG. 1, the basic configuration of the video conference apparatus of the present invention comprises an image input means 1, a distance detection means 2, an image processing means 3 and a display means 4. The distance information detected by the distance detection unit 2 is added to the image information input by the image input unit 1, only the image of the necessary portion is extracted by the image processing unit 3, and the image is enlarged and displayed by the display unit 4. It is projected. In FIG. 2, 5 is a main TV camera, 6 is a sub TV camera, 7 is a zoom-up TV camera, 8 is an entire display TV monitor, 9 is a speaker display TV monitor, 10 is a moving body discrimination means, and 11 is a distance calculation means. , 12 is speaker determination means, 13 is image enlargement means, 14
Is composed of a camera controller. The operation will be described based on FIG. The input image is displayed as it is on the entire display TV monitor 8 on the main TV camera 5 arranged so as to display the conference participants. The input image obtained by the main TV camera 5 is sent to the moving body discriminating means 10 to discriminate the position of each conference participant. Further, the sub television camera 6 is arranged in parallel with the main television camera 5, the image of the main television camera 5 and the image of the sub television camera 6 are compared, and the distance from the television camera to the subject is calculated by using the trigonometry. It is calculated by the calculation means 11. It is preliminarily decided that the conference participants should raise their hands before speaking. When the person who is going to speak takes some action, the images of the main television camera 5 and the sub television camera 6 change. Therefore, the speaker is identified from among the participants recognized by the moving body determination unit 10 based on the change in the image. The position information determined by the speaker discrimination means 12 is sent to the camera control device 14,
The zoom-up TV camera 7 is adjusted to the direction of the speaker. At this time. By sending the distance information calculated by the distance calculation means 11 to the camera control device 14, the focus of the zoom-in television camera 7 can be started even while the camera orientation is being adjusted. When the zoom-up TV camera 7 completely captures the speaker, the image on the speaker display TV monitor 9 is switched from the image of the image enlarging means 13 to the screen of the zoom-up TV camera 7. At this time, a natural image switching is performed by matching the image of the zoom-up TV camera 7 with the image created by the image enlarging means 13 and framing the zoom-up TV camera 7 so as to obtain the most similar image. Is possible. In FIG. 3, (a) is a schematic diagram showing the positional relationship between the television camera and the subject, (b) is a diagram showing an image obtained from the main television camera 5, and (c) is a sub television camera. 6
It is the figure which showed the image obtained from. Main TV camera 5
It is assumed that the sub television camera 6 and the sub television camera 6 are arranged in parallel with each other, and the subject is located just in front of the main television camera 5. The angle of view of both the main TV camera 5 and the sub TV camera 6 is 2θ, and the main TV camera 5 and the sub TV camera 6
When the distance between and is D and the distance between the main TV camera 5 and the subject is 1, the image set by the TV camera on the straight line L is within a range of ± 1 tan θ with respect to the front, and the image set by the TV camera and the sub-TV are set. If the same subject as the image set by the camera 6 is deviated by d, this d is d
= D / 1 tan θ. That is, 1 = D / dta
nθ, and the distance 1 from the television camera to the subject is calculated from the shift d on the screen. A block matching method or the like can be used when obtaining the deviation between the images of the subject between the image of the main TV camera 5 and the image of the sub TV camera 6. From the above, the depth corresponding to each I pixel of the screen is obtained, and the obtained value will be referred to as the z value hereinafter. When the speaker raises his / her hand by the speaker determination means 12 shown in FIG. 2, the arm of the speaker comes to the rear wall or to the front of the speaker's own body, so that the z value of the portion that hits the arm on the screen decreases. Will be done. The speaker is determined based on this decrease in z value. Further, it is possible to use a motion vector as a means for confirming the motion of the raised hand. People usually seem to have little movement during a meeting. Among them, the motion of raising hands is relatively large,
It is a fast movement. This can be represented by a motion vector extracted from between two frames, and the speaker can be recognized from the magnitude and speed of the motion. In FIG. 4, linear interpolation can be considered as a simple image enlarging means. When trying to enlarge an image, a gap is created by simply increasing the distance between pixels. Linear interpolation is performed to fill this gap. The value of the pixel to be interpolated is calculated and calculated from the surrounding four neighbors. The point to be interpolated is E, its value is e, 4 neighboring points are A, B, C, D, its value is a, b,
Let c and d. The distance between pixels before enlargement is 1, point E is point A
Is x in the horizontal direction and y in the vertical direction, the value e of E is: e = (1-x) (1-y) a + x (1-y) b + (1-
x) yc + xyz.

This makes it possible to create a pseudo zoomed-in image.

[0009]

As described above, according to the present invention, a zoom-up TV camera for tracking a speaker in a video conference cuts out the entire image and enlarges the speaker image in advance even when the orientation is adjusted. By displaying the image, it is possible to quickly send the image of the speaker to the other party of the video conference.

[Brief description of drawings]

FIG. 1 is a block diagram showing a basic configuration of a video conference device of the present invention.

FIG. 2 is a block diagram of a video conference device of the present invention.

FIG. 3 is a diagram illustrating a distance calculating unit of the video conference apparatus of the present invention.

FIG. 4 is a diagram illustrating an image enlarging unit of the video conference apparatus of the present invention.

[Explanation of symbols]

 1 Image Input Means 2 Distance Detection Means 3 Image Processing Means 4 Display Means 5 Main TV Camera 6 Sub TV Camera 7 Zoom Camera 8 Whole Display Monitor 9 Speaker Display Television Monitor 10 Moving Object Discriminating Means

Claims (1)

[Claims] 1. A video conferencing apparatus having display means for displaying an object photographed by a camera, comprising: an operation judging means for judging an operation of the object; and an extracting means for extracting the object judged by the operation judging means. An image enlarging means for enlarging the image of the subject extracted by the extracting means, a photographing means for photographing the image enlarged by the image enlarging means, and a display means for displaying the image photographed by the photographing means. A video conferencing device characterized by being configured.