JPS6187489A - Television conference system - Google Patents

Abstract

PURPOSE:To reduce a conference space by detecting a talker based on a voice signal from a microphone for TV conference and picking up the image of talker only with installation of a TV camera for human figure so as to save number of cameras. CONSTITUTION:When the number of talker detected by a talker detection circuit 22 is one, a turning base control servo controller 28 and a lens control servo controller 30 are operated. Then a position data of a channel number corresponding to the detected talker is read, a turning base 16 and a lens 18 are controlled based on the read position data to direct the TV camera 12 for human figure to the talker corresponding to the said channel. A selection command of the camera 12 is given to a video switching circuit 23 from the video control section 24c after a prescribed delay time. Thus, the talker in a TV conference is picked up automatically and the video signal of the camera 12 is transmitted to the opposite office via a video transmitter 34.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a video conferencing system, and more particularly to an improvement in a system for connecting remote locations using audio and video transmission to conduct a teleconference.

[Prior art] A video conferencing system connects a long distance to an office using eight voices and a video, and allows attendees to be seen on a screen at a meeting in Aiko's office (while all attendees are in the same room). It is possible to conduct a meeting in a friendly atmosphere.

Therefore, by using such a video conferencing system in Song Dynasty and holding a long-distance meeting between fXI and A rices;
It is expected that there will be a wide range of demands in the future, including the ability to save money on business trips and, in particular, to quickly hold emergency meetings.

[Problems to be Solved by the Invention] By the way, such a video conference system generally includes a person TV camera that records the speaker, a panoramic TV camera that displays the names of participants in the video conference, and a video camera that displays the names of participants in the video conference. In particular, a plurality of TV cameras for people are installed so that they can take decoy images of the speakers individually, and each TV camera for each person is installed facing the seat of the attendee. was.

Therefore, in conventional video conferencing systems, the number of person-viewing television cameras used is constantly increasing, and the entire system becomes extremely large and has a large α value.

Furthermore, in conventional video conferencing systems, the moderator or assistant manually controlled the switching of the television cameras for each person and the metal television cameras, so the operator had to switch the television cameras every time the speaker changed. Switching control must be performed, which makes the operation extremely complicated and has the drawback of being prone to operational errors.

Purpose of the Invention The present invention has been made in view of such conventional problems, and its purpose is to make it possible to automatically photograph each speaker with a small number of television cameras for people. An object of the present invention is to provide a video conference system that can automatically control switching between a camera and a medical TV camera.

[Means and effects for solving the invention] In order to achieve the above-mentioned object, the video conference system of the present invention [Yo, the installation position data of Sugan's microphone installed for video conferences is stored in advance in the F-tameter memory. The caller 58 detecting circuit detects the sender 6 based on the 8th almond number from the 3rd fix. The control circuit then controls the position 1-ta memory/).
The position data of the microphone used by the speaker is read out from s+, and based on this read position data, the person television camera is controlled to perform the dance of the speaker.

Therefore, according to the present invention, it is possible to completely automatically photograph a speaker in a video conference using at least t) one television camera for people. .

In addition, other inventions of the present application [by the video & switching circuit,
If there is only one speaker detected by the utterance circuit '8', the video signal from the human TV camera is selectively output, and if the speaker is cold, the video signal from the panoramic TV camera is selectively output. There is.

In this way, it is possible to automatically select and switch between a pre-camera for people and a video camera depending on the number of participants (J, S, 8, N), and select the optimal image signal.

[Example] Next, preferred embodiments of the present invention will be described based on the drawings.

(A) Configuration Figure 2 shows the basic configuration of the system for a TV conference. A conference device 110 and a transmitting/receiving terminal 120 are installed in each office, and the two offices 100A and 100B are connected by audio and video.

Therefore, the participants in the meeting at each office 100A and 100B can proceed with the meeting while watching the notary participants in the other party's office through the screen, so the meeting can be held with the atmosphere that everyone is in the same room. It can be carried out.

Figure 1 shows that in such a video conference system,
A circuit is shown for creating an image of attendees in one office.

The video conference system of the embodiment includes a plurality of microphones 10-1, 10-2, and 10-2 for video conferences in a conference room of an office 100.
...and TV camera 12 for people) A TV camera 14 for panoramic view is provided.

Here, the prefecture-wide television camera 14 is used for a television conference;
Participant names are to be photographed within the whole, and accordingly,
The camera body is fixed so that it can always display a constant image. Although the system is set up to take pictures of all the participants, it is also possible to omit this and, for example, use two cameras to take pictures of half of the participants.

Further, the television camera 12 for persons is configured to individually record eight persons speaking to each microphone, and therefore,
The television camera 12 selects a person to be photographed by rotating a rotary table 16, and adjusts the perspective of the person to be photographed by adjusting a lens 18.

That is, it consists of two-axis adjustment in the X and Y horizontal directions by the rotary table 16, and a Z@ adjustment representing perspective on the lens 18.
The three axes of X, Y, and Z are adjusted.

The first characteristic feature of the present invention is to automatically detect speakers in such a video conference, and use the television camera 12 for people to record yi! The goal is to automatically take pictures of speakers.

Therefore, in the system of the present invention, each microphone 10
-1.10-2... position data memory 20 in which installation position data is stored in advance, and each microphone 10-1.
It includes a speaker detection circuit 22 that detects a speaker based on the audio signals from 10-2, . . . .

In the embodiment, the position data memory 20 stores position data of each microphone 10-1, 10-2, etc. (
The information is stored as three-axis information (X, Y, Z).

In the embodiment, the speaker detection circuit 22 detects the user of the microphone from the microphone 10-1. 14, and supplies the Brenner pump of the microphone to the control panel 24 and control circuit 26.

Here, since the pitch of the speech and the speaking time of Akane, who is attending the meeting, are different from each other, the system rlA of the embodiment recognizes it as the voice signal of the speaker by the reference level setting unit 24a of the control group 24. Standard tone? The level and speaking duration level for recognition as a speaker are set. Then, the speaker detection circuit 22 detects the speaker based on the reference volume level and reference speech duration level set for each microphone.

When the control circuit 26 receives the block number of the microphone 10 used by the detected speaker from the speaker detection circuit 22 in this way, the control circuit 26 reads the block number of the microphone from the position data memory 20.・Do you agree? Read out the mounting data and adjust the rotary table 16 based on this position data.
Then, the lens 18 is controlled and the person television camera 12 is used to photograph the RE.

In the embodiment, this control circuit 26 controls the
, -Y control servo 1"r'J
- La 28 and I, i”i; 1.16 (7) Z@, Suj
;b'5M It consists of a servo controller 30 for lens control that performs near adjustment.

The rotary table control servo controller 28 reads out the X and Y data of the corresponding microphone from the position data memory 30 based on the channel number inputted from the speaker detection circuit 22, and controls the pitch and water of the rotary table 16.

Similarly, the lens control controller 30 retrieves data in the two-axis directions of the corresponding microphone 10 from the position data memory 20 based on the channel number input from the speaker detection circuit 22. Readout and servo llJ control of the lens 18 are performed.

In this way, according to the video conference system of the present invention, the person television camera 12 is controlled based on the audio signals from each of the microphones 10-1, 10-2, etc., and the person speaking in the video conference is photographed. It can be done automatically.

In the system of the embodiment, the control panel 24 is provided with a camera control section 24b that switches the human television camera 12 between automatic and manual switching so that the human television camera 12 can be manually controlled. By setting to the manual side, the servo control roller 28 for controlling the rotating table and the servo controller 1-30 for controlling the lens 11 can be manually controlled by the joystick, and any image can be captured by the television camera 12 for people. can be done.

By the way, when photographing the speaker in such a video conference, if there is only one person, the TV camera 1 for the person is used.
It is preferable to obtain an image of the speaker using the TV camera 14, but in the case where multiple people are watching at the same time, it is preferable to use the panoramic TV camera 14 to obtain an image of the entire speaker in the video conference.
7 is preferred.

The second characteristic feature of the present invention is the television camera 12 for people.
A video switching circuit 32 is provided at the output stage of the television camera 14 for people. The purpose is to selectively output the video signal, and when the number of speakers detected by the speaker detection circuit 22 is plural, the video signal from the whole-body television camera 14 is selectively output.

In the actual tMVA, this video switching circuit 32 connects the video controller 1 to the roll section 24c provided in the control I panel 24.
Video signals from each camera 12 and 140 are selectively output in accordance with a switching command from the camera.

Video control unit 24 of control 1 pIA 24. c is provided with a video signal auto/manual I, U changeover switch, and by setting the on switch to the auto side, this video control section 24 controls the output of the three sources detected by the Ga emission detection circuit 22. Based on the number, the video signal switching circuit 32 automatically outputs video switching signals. This video switching signal instructs the TV camera 12 to selectively output the video Nobuoka when the speaker detected by the speaker detection circuit 22 is -IJ.
If there is a speaker, or if there is no speaker, the system instructs the panoramic television camera 14 to selectively output the image I&signal.

Further, by setting the switch of the video control section 24C to the manual side, it becomes possible to manually select and output the image of the person television camera 12 or the panoramic television camera 14 as desired.

In this way, the video l! selectively output from the video switching circuit 32! The signal is transmitted to the other party's office by the video transmission device 34.

(B) Effect This example is the above 41? Next, the operation of this embodiment will be explained based on the flowchart shown in FIG.

Camera controller on control panel 24!・] When the off switch of the call section 24b and the video control section 24C is set to the O1~ side (steps 200 and 210),
The video switching circuit 32 is a video control 81!24c
In response to a command from , an initial setting is made to selectively output the video signal of the panoramic television camera 14 .

Here, an arbitrary microphone, for example, microphone 10-1
Assuming that the number 8 Ann Jin is input from channel CH1 connected to step 310), and if the volume level has increased, steps 320 to 360 and steps 400 to 430 indicate
Perform a series of speaker detection operations.

Incidentally, as described above, the reference sound level set by the reference level setting section 24a? The level can be set to an arbitrary value for each channel, and the speaker detection circuit 22 uses the reference volume level set for each channel in this way to determine the level of each channel. The volume level of the channel audio signal is determined.

In this way, when the three-caller detection operation for any channel CH1 in this embodiment is started,
The speaker detection circuit 42 determines whether or not an audio signal is continuously input from CH1 for a certain period of time, and when the audio signal is continuously input for a certain period of time, Identify the corresponding person as the speaker.

That is, at the same time as the speaker detection circuits 2, 2 start the speaker detection operation, the tie? The first counter is incremented by 1 (step 330), and the counter is incremented by 1 when it is determined that the volume of the audio signal has exceeded the reference total level (step 330). 350). Here, the detection cycle of each of steps 330 to 350 is repeated every second, and this count is the IQ' count corresponding to the reference oscillation Δ duration level set by the jit type level setting section 24a. When reaching (step 350), a speaker detection signal for channel C111 is output (step 360). 1 At the same time, steps 330~
The operation consisting of 350 is repeated.

It should be noted that steps 330 to 350 are performed in step 3. The Ministry of Detection Standard is J! where the audio signal from the microphone 10 is constant. This shows the operation when the quasi-utterance duration is input continuously, but in reality, the speaker speaks while thinking about the content of the light 8, so
The audio signal from the microphone 10 is often input discontinuously and continuously for 1fJi.

Therefore, in this embodiment, even when the audio signal is input intermittently, as long as the time interval between the high-angle signal and the audio signal is within a certain reference time, this signal is input as a series of continuous signals. It is recognized as a statement and an operation to detect the speaker is performed (steps 400 to 430).

For example, if it is determined in step 310 that the audio signal from each channel 11 is equal to or higher than the sound R/fi reference level, but the audio signal is subsequently interrupted (step 330), the interruption time of this audio signal is In order to detect this, a second counter as a timer is reset (step 400), and this audio interruption time is detected every second, and the counter is incremented by one every second (steps 410 to 430). ).

If an audio signal of a level equal to or higher than the reference B% level is input from the channel C111 until the count of the second counter reaches the reference number 1 corresponding to a predetermined reference time, the speaker detection circuit 22 The operation is step 34
The value returns to 0, and the series of utterance 1'i detection operations described above (steps 330 to 360) are continued.

However, no audio signal is input from channel CHI,
When the count of the second counter exceeds the reference count (step 430), the speaker detection circuit 22 determines that the transmission of channel CH1 has been stopped, and returns to step 300.

The above is the detection operation of the f-channel CHI, and the present invention performs such detection for each of the 11 channels.

Then, the speaker detection circuit 22 detects each microphone 101 .
Based on the audio signals from each channel corresponding to 10-2..., one speaker of the video conference is announced.

At this time, the present invention detects the speaker depending on whether or not the audio signal corresponding to each channel is continuously input for a certain period of time. It is possible to prevent malfunctions such as recognizing a person's name as a utterance and to reliably detect an Ffi speaker.

Operation of TV camera for people Here, if the number of speakers detected by the three-person speaker detection circuit 22 is 1, that is, if the number of speakers is -. 28 and the lens control turbo controller 1-[Cola 30 operates, reads the position data of the chilennel pumper corresponding to the detected utterance B, and based on the read position data, the rotary table 16.
Control the lens 18 and set the TV camera 12 for people to the three people corresponding to the camera (step 5).
00).

It should be noted that it takes a certain amount of time to mechanically move the camera 12 to set up the TV camera 12 for people. After a predetermined five hours have elapsed, a command to select the human television camera 12 is output to the video switching circuit 32.

In this manner, according to the present invention, the Ud shadow of the speaker in the video conference is automatically performed, and the video g of the TV camera 12 for the person is directed to the other party's office via the video transmission device 34. It will be sent to you.

In addition, in step 500, the F+C detection circuit 22
If the number of speakers detected by N/) < N≠1,
In other words, if there are multiple pronunciation speakers at the same time or
If the name is unknown, a command to select the personal television camera 14 is supplied from the video controller 1 to the roll unit 24C to the video switching circuit 32 (step 700).

As a result, the video trimming circuit 32 selectively outputs the video No. 45 of the panoramic television camera 14 and sends it to the other party's office via the video transmission device 34.
The video of the video will be transmitted.

In this way, according to the present invention, depending on the number of speakers, the person television camera 12 and the panoramic television camera 1/l
4 can be automatically selected and used, and the video signal can be transmitted to the other party.

Although the above description has been made by taking as an example the case where a video signal is transmitted as a moving image via the video transmission device 34, the present invention is not limited to this, and the present invention is not limited to this. It is also possible to transmit the image to the other party's office as a still image that is updated at regular intervals, thereby reducing the number of transmission lines used.

Steps 630 to 660 and steps 720 to 750 show operations in the case of transmitting such still images, and steps 630 to 660 specify the video order number of the television camera 12 for people. The operation in the case of transmitting at each time interval is shown, and steps 720 to 750 show the supplementary operation in the case of transmitting the video signal of the panoramic television camera 14 at each predetermined time interval.

In other words, the image 1τ of the human television camera 12:
When the selected IR F'L is selected (step 620), when transmitting such a still picture at a predetermined interval of 7, first reset the 30th counter as a timer (step 630). , the content of the third counter is set to 1 each time (step 640), the content of the third counter is set to 1.
The third counter is then incremented (step 650), and when the content of the third counter exceeds a predetermined reference count, the next video signal of the human television camera '12 is read (step 620).

In this way, the image number 1g of the person's reference camera 12, that is, the still image, which is updated at every predetermined 1.1 interval set by the reference count, is transmitted to the i-th Teno's office.

In addition, in the same manner as steps 620 to 660, a fourth counter as a timer is set every time the video signal of (ma) is output to the platform on which the all-day TV camera 14 is selected (step 700). The contents of the fourth non-counter are incremented every second by resetting the contents of 1 to N゛・1. When I went around Motokan Count 1-, I saw the footage of the next Zenroyo TV camera 14 of No. 1 (Ici).
Performs a program transfer.

In this way, the video signal of the gold seal TV camera 14 is transmitted as a still image to the other party's office at a predetermined interval.

In addition, when performing the transmission after such a still image, when performing the operation shown in step 640, if it is determined that the number of detected speakers N is not one, the video switching circuit 32 The video signal to be selectively output is controlled to be switched to 14 signals of the metal television camera. Similarly, step 73
0! During the process of FJJ V+ 'i, if it is determined that the number of detected issuers is not N≠1, the video switching circuit 32 changes the number to a predetermined value of +1. ”J period elapsed v2 (step 610)
, the video signal of the person television camera 12 is switched to be selectively output.

Note that these steps 630-660, 720-750
It is necessary to set the reference counts of the third and fourth counters shown in the steps 630 to 660.720 to values corresponding to the modem speed of the transmission system.
When transmitting a moving image using a device in which .about.750 is crudely manufactured, it is sufficient to set the reference number 1 to 0 to O.

? 21 AL OPERATION Also, in this embodiment, there is a TV camera 12 for people.
or video (<9 t, when trying to manually operate the video signal selection switch of the IJ switching circuit 32, the camera control section 24b or video controller 1~) 1 - 2
/4C selector switch can be set manually.

[Effects of the Invention] As explained above, according to the present invention, it is possible to detect a caller based on a high-pitched signal from a microphone installed for a conference. Since it is possible to perform a dance with the speaker by simply installing a TV camera for one power, it is possible to reduce the number of TV cameras used and reduce the if of the meeting space compared to the second one. In addition, in the present invention, since it is not possible to automatically carry out the preparation of each speaker, it is possible to avoid the trouble of switching the television camera and smoothly conduct the preliminary stage of the meeting. That's possible.

Furthermore, according to the present invention, it is possible to automatically select and use a TV camera for people and a dual-purpose camera depending on the number of speakers in a video conference. It is possible to save the time and effort required to switch to a TV camera for all threats, and to proceed with the meeting even more smoothly.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a preferred embodiment of the video conference system according to the present invention, FIG. 2 is an explanatory diagram showing the general configuration of the video conference system, and FIG. 3 is the operation of the system shown in FIG. 1. It is a flowchart figure which shows. 10...Microphone 12...Television camera for people 14...Television camera for full volume 20...Position data memory 22...Speaker detection circuit 26...Control circuit 30...Video switching circuit No. Figure 1

Claims (5)

[Claims] (1) A plurality of microphones installed for video conferencing, an entrance television camera that individually photographs speakers speaking to each of the microphones, and positions where the installation position data of each of the microphones is stored in advance. a data memory; a speaker detection circuit that detects a speaker based on the audio signals from each of the microphones; and a speaker detection circuit that reads position data of the microphone corresponding to the detected speaker from the position data memory and detects the person's television based on the read position data. A television conference system comprising: a control circuit that controls a camera to photograph a speaker, and automatically photographs a speaker in a television conference. (2) The system according to claim (1), wherein the speaker detection circuit detects a microphone that continuously outputs an audio signal for a certain period of time as a microphone used by a speaker. . (3) Multiple microphones installed for video conferences, an entrance TV camera that individually photographs the speakers speaking to each of the microphones, and a panoramic view that photographs the names of participants in the video conference. a position data memory in which installation position data of each of the microphones is stored in advance; a speaker detection circuit that detects a speaker based on audio signals from each of the microphones; a control circuit that reads position data from a position data memory and controls a TV camera for persons to photograph a speaker based on the read position data; A video switching circuit that selectively outputs the video signal of the television camera and, in the case of multiple speakers, selectively outputs the video signal of the panoramic television camera, and automatically switches between the human television camera and the panoramic television camera depending on the number of speakers. A video conference system characterized by selective use. (4) The system according to claim (3), wherein the speaker detection circuit detects a microphone that continuously outputs an audio signal for a certain period of time as a microphone used by a speaker. . (5) In the system according to any one of claims (3) and (4), the video switching circuit switches the panoramic television camera when the number of speakers detected by the speaker detection circuit is zero or more than one. A video conference system characterized by selectively outputting video signals.