JP2708480B2 - Video teleconference equipment - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Summary] A transmission image switching control in a video conference of a method of transmitting a specific subscriber image among the subscriber images of a conference to all the participating subscribers within a predetermined period. For video teleconferencing equipment using utterance probability, multiple voice CODECs for video teleconferencing aiming to transmit an appropriate image source to each participant and prevent image quality degradation due to excessive switching of image sources. And a video telephone conference device having an output image code selector, an arithmetic unit for outputting an average audio power value for each average power determination period for each participating subscriber, and each audio power average value is predetermined. A determination unit that determines whether the threshold value is exceeded, and a predetermined multiple of the average power determination period for the determination in response to the determination output corresponding to each participating subscriber output from the determination unit. A determination unit for determining one participating subscriber whose utterance probability is equal to or greater than a predetermined value for each voice probability calculation period, and the output of the determination unit is used to cause the output image code selector to perform a selection operation. Was configured.

[Industrial applications]

The present invention relates to a videophone conference device that uses utterance probabilities within a predetermined period for image switching control in a videoconference.

An ISDN network that enables faster information transmission than a conventional analog telephone network has been put into practical use. On the other hand, there is an increasing demand for a video teleconference service that is more realistic than a conventional audio conference call. Therefore, provision of a video conference system utilizing ISDN effectively has been studied and is being put to practical use.

[Conventional technology]

The conventional video conference system is shown in FIG. In this figure, 100 ₁ ,..., 100 ₄ are video telephone terminal devices for each subscriber, 102 is a digital exchange, and 104 is a video telephone conference device. 106 ₁ , 108 ₁ , 106 ₂ , 108 ₂ , ..., 10
6 _4, 108 _4, respectively, are stretched when the call between the corresponding videophone terminal and the digital switch 102, for example 2
These are the B1 channel and the B2 channel, respectively, of the IS + channels of the B + D configuration. 110 ₁ , 112 ₁ , 110 ₂ , 112 ₂ , ..., 11
0 ₄ and 112 ₄ are the digital exchange 102 and the video conference unit 10
4 are the B1 channel and the B2 channel of the same ISDN channel formed as above. The B1 channel and B2 channel are used for transmitting and receiving audio and moving images as shown in the figure.

When each of the illustrated subscribers participates in the videophone conference, the above-mentioned channels are connected to the respective videophone terminal devices 100i (i = 1, 2, 3, 4) and the digital exchange 102 under the conventional communication channel setting method. And between the digital exchange 102 and the videophone conference device 104. Then, voices and moving images as shown in the figure are transmitted from the video telephone conference device 104 to each subscriber terminal device via the digital exchange 102 via these channels, and the video telephone conference proceeds.

[Problems to be solved by the invention]

For the audio signal in the above-described video conference call system, even if the technique used in the conventional audio conference device is directly used in the video conference device, for example, 64K
It is considered that there is no problem in speech quality even if a signal of the PCM coding system of bits / second is received, and voice signals of the participating subscribers are linearly added, coded and returned to each subscriber. This is because the voice signal of each subscriber in the video teleconferencing system is such that the subscribers other than the speaker are usually in a silent state and are in a burst-like state. This is because it is considered that no trouble occurs.

On the other hand, the image signal must continuously supply the required image signal to each participant independently of the speech. If an attempt is made to transmit the image signal using the same band as the audio signal (B channel of ISDN), the image signal having a signal band wider than the audio signal band must be compressed without a significant signal band compression. Cannot be transmitted.

After decoding such band-compressed signals and image signals from each participating subscriber in the same manner as audio signals, and linearly adding them,
If the addition signal is coded and transmitted, the image quality is degraded in proportion to the increase in the amount of information to be coded in proportion to the number of participating subscribers.

In order to avoid this, an image signal of only the participating subscriber who is speaking is distributed to all other participating subscribers, and a voice signal may be a conventional video telephone conference system. It is conceivable to use audio signal power as a distribution control method of the image signal. For example, there is a method in which the audio signal power of each participating subscriber is compared, and the image signal of the participating subscriber indicating the maximum is distributed to all the participating subscribers.

If the image signal from each participant that has been encoded by the high-efficiency encoding method that encodes and transmits the difference component between frames is to be switched in the above-described method, when the switching is performed, first, Since it is necessary to distribute the original image data for one frame for obtaining the difference component to each participating subscriber without performing the encoding, the image quality deteriorates during that period. Therefore, the image of the participating subscribers in a situation where the switching frequently occurs is not stable, and the image quality may be extremely deteriorated.

The present invention has been created in view of such problems,
It is an object of the present invention to provide a videophone conference device that prevents image quality degradation due to an incoming call of an image according to a speaker to each participating subscriber and excessive switching of images.

[Means for solving the problem]

FIG. 1 is a block diagram showing the principle of the present invention. In this figure, i ₁ (i = 1, 2,..., N) is one information channel in the ISDN channel corresponding to each subscriber, and is a channel used for an audio signal for a video conference call. (Hereinafter abbreviated as ISDN voice channels), and i ₂ is another information channel in each ISDN channel, which is a channel for image signals (hereinafter abbreviated as ISDN image channel). Each. 2 _i is each of audio CODECs provided for each ISDN audio channel. 4 is each voice C
An arithmetic unit connected to the output of the ODEC2 _i and for outputting an average audio power value during an average power determination period for each participating subscriber. Reference numeral 6 denotes a determination unit that determines whether or not each audio power average value exceeds a predetermined threshold. Numeral 8 indicates that the utterance probability is equal to or greater than a predetermined value in each utterance probability calculation period that is a predetermined multiple of the average power determination period for the determination in response to the determination output corresponding to each participating subscriber output from the determination unit 6. Is a determining unit for determining one participating subscriber. The present invention is configured so that the output of the determination unit 8 causes the selection operation of the output image code selector 10.

[Action]

The average value of the audio power of each subscriber participating in the video conference is output from the arithmetic unit 4 every average power determination period. A determination output as to whether or not the average value of the audio power exceeds a predetermined threshold is output from the determination unit 6. The determination output corresponding to each participant within the utterance probability calculation period is used by the determination unit 8, and one participant whose utterance probability is equal to or more than a predetermined value is determined. The image code of the determined participant is output to the output image code selector.
Output from 10. When the utterance probability of any of the participating subscribers is less than the predetermined value, the determination is not made, and the determination in the past utterance probability calculation period is used for the selection operation of the output image code selector 10.

〔Example〕

FIG. 2 shows an embodiment of the present invention. In this figure,
2 _i (i = 1, 2,..., N) are audio CODECs and each audio C
ODEC is a voice channel for videoconferencing of digital exchanges (one B channel of ISDN channel of 2B + D configuration)
It is connected to the i _1. Reference numeral 10 denotes an image channel for a video conference of a digital exchange (the other B of the above ISDN channel).
An output image code selector for sending by folding the image data came transmitted through one image channel of the channels) i ₂ for all image channels. The selection operation of the selector 10 is caused by a selection signal transmitted via the line 30.

The output of the decoder 12 _i of each audio CODEC 2 _i is an echo canceller
The 16 _i filter 18 _i is connected to the adder 24 and the transmission image control circuit 26. The output of the adder 24 is each echo canceller
16 _i is connected as a subtracted input to the subtractor 20 _i
i output is connected to the encoder 14 _i of the corresponding voice CODEC2 _i. The adder 24 is composed of (n-1) adder circuits 24 ₁ , 24 ₂ ,.
- consists of 24 _n-1, adding circuit 24 ₁ is the sum of the decoder 12 ₁ and 12 ₂ of the output signal, the output signal of the addition circuit 24 ₂ addition circuit 24 ₁ of the output signal and the decoder 12 ₃ and Is added. Addition circuit 2
Summing the input signal to be from 4 ₃ to the addition circuit 24 _n-1 is the output signal of a decoder indexed by adder circuit and one number greater than indices indexed by one younger number than its index. Further, the subtraction input to the subtracter 20 _i is the output signal of the corresponding filter 18 _i.

Transmitting the image control circuit 26, the decoder 12 _i corresponding speech power mean value calculating circuit 40 _i (see FIG. 3), the comparator 42 _i which receives the output of the audio power average calculation circuit 40 _i, the comparator 42 _i All adds _i counter 44 which receives the output count input, 1-bit left shift register 46 _i receiving the count output value in each counter corresponding, all counters 44 _1, 44 _2, ..., the count output value of 44 _n The adder 48 is a comparator 50 _i corresponding to each 1-bit left shift register for comparing the shift output value with the output value of the full adder 48, and receives the binary output of each comparator 50 _i at the set input. In addition to the register 52, an 8-bit counter 41, a 10-bit counter 43, an OR circuit 45, and an AND circuit 47 are provided. 8KHz count input of 8 bit counter 41
Although the system clock is supplied, the clock is also supplied to each of the voice power averaging circuit 40 _i as the audio power average calculation circuit 40 _i each register clock (described later). 8 carry output of the bit counter 41 (divided output) with is supplied to the clock input of counter 44 _i, are supplied to the count input of a 10-bit counter 43. Carry output of the 10-bit counter 43 is supplied to one input of a reset input and the AND circuit 47 of each counter 44 _i. The other input of the AND gate 47, the output of the OR circuit 45 which receives the binary output signals from the comparators 50 _i are supplied. The output of the AND circuit 47 is supplied to the clock input CK of the register 52.

The audio power average value calculation circuit 40 _i corresponds to the calculation unit 4 in FIG. 1, and the comparator 42 _i corresponds to the determination unit 6. 8-bit counter 41,10-bit counter 43, the counter 44 _i, 1
Bit left shift register 46 _i , full adder 48, comparator 50 _i ,
The OR circuit 45, the AND circuit 47, and the register 50 correspond to the determining unit in FIG.

The operation of transmission image control according to the embodiment of the present invention will be described below.

In the embodiment of the present invention, an initial screen is shown on the screen of each videophone terminal (participant) when the videophone conference system starts operating. Then, an image of a moderator or one of the subscribers who is uttered via the video conference call apparatus in response to the utterance made at the start of the video conference call and which is determined as described later is displayed. It is configured to be shown to other participants.

 The switching of the screen will be described below.

Also in the embodiment of the present invention, as in the prior art, each subscriber is sent a voice signal from the video teleconference apparatus, voice signals of all other subscribers except for the subscriber are transmitted, and images are spoken during the conference. An image of one of the participants, determined as described below, is sent to all participants in the conference. The audio signal system for this is the audio shown in Fig. 2.
CODEC2 ₁ , 2 ₂ ,..., 2 _n , echo cancellers 16 ₁ , 16 ₂ ,
.., 16 _n and an adder 24. That is, an audio signal obtained by adding the audio signals of all the subscribers participating in the conference is output from the adder 24, and the echo canceller 16 _j (j is the number of all the participants participating in the conference) from the synthesized audio signal. of representing each of the subscriber.) synthesized speech signal which remains clear the audio signal of the corresponding subscriber is encoded in the encoder 14 _j in, through the voice for B-channel j ₁ of the subscriber corresponding ISDN channel j It is sent back to the corresponding subscriber _Sj . The elimination of the self-voice signal in the echo canceller 16 _j is for preventing the occurrence of howling or the like by solving the formation of the closed loop. A selection signal to the output image code selector 10, which selects one of the image data to be sent to each subscriber from among the participating subscriber image data and outputs it, is transferred from the transmission image control circuit 26 via the line 30. The method of generating the selection signal will be described below.

Each subscriber signal power mean value for S _j of the audio signal (the output signal of the decoder 12 _j) (root mean square) of Figure 4 (but without adding the subscripts j for participation subscriber in FIG _Is output from the audio power average value calculation circuit _40j shown in FIG. That is,
The output value of the multiplier 66 _j one clock (8 KHz) earlier and the output signal of the decoder 12 _{j at the} average output clock time and the signal from the squarer 60 added by the adder 62 _j are used as the average output time. The above-described average value of the digital signal power is output by setting the clock in the register _64j and outputting it. The adder 62 _j , the register 64 _j and the multiplier 66 _j constitute a leak integrator.

Each digital signal power average value is compared with a digital threshold value (preset value) by a corresponding comparator. When the signal power average value exceeds the threshold value, the comparator outputs "1".
Is not exceeded, “0” is output. The comparator output "1" is counted by the corresponding counter every average power determination period (for example, 32 msec in the above clock example) set by the 8-bit counter 41. The counting is performed during an utterance probability calculation period (for example, 32 seconds in the above-described clock example) set by the 8-bit counter 41 and the 10-bit counter 43, and when the period elapses, one of the counters _44j is reset. The count value of each counter 44 _j is summed by the full adder 48 for each count and doubled in the corresponding 1-bit shift circuit. When the output value of the 1-bit shift circuit exceeds the output value of the full adder 48, the corresponding comparator 50
An output of “1” is generated from _k (k is any of j), and otherwise, “0” is output.

Outputting “1” from only one of the comparators 50 _k means that the utterance probability of the subscriber corresponding to the output is 0.5 or more, and the image of that subscriber is transmitted to each participating subscriber. Is done. That is, the output of "1" is given to the AND circuit 47 via the OR circuit 45, and when the carry output from the 10-bit counter 43 is generated, the clock is given from the AND circuit 47 to the register 52 to register the "1". Set to 52 corresponding bit positions. The set output (transmission right for image transmission for the corresponding participant) is supplied to a selection input of the output image code selector 10 via a line 30, and the corresponding image data is transmitted via the B2 channel of the participant ISDN channel. Distributed to each subscriber as described above.

Also, for each utterance probability calculation period, when utterance probability described above that time it is less than 0.5, the output of either comparator 50 _j also "0" at the contents of register 52 updates, i.e. to participants subscriber Switching of the transmission image (movement of the transmission right) does not occur. As a result, the switching frequency of each subscriber screen can be reduced and the sense of incongruity given to participating subscribers can be reduced.

In the above embodiment, an example in which the present invention is configured by providing circuits for calculating, comparing, counting, shifting left, and comparing audio power average values for each ISDN channel has been described. As long as no inconvenience occurs, a common circuit, for example, a microprocessor may be used for each required number of ISDN channels. The utterance probability may be set to a value other than 0.5.

〔The invention's effect〕

As described above, according to the present invention, the utterance probability at the time when the average utterance power of each participating subscriber is obtained using a frequency exceeding a predetermined threshold within a predetermined period is obtained. Since the transmission right is determined depending on whether or not the value exceeds a predetermined value, frequent switching of images transmitted to the video conference participating subscribers is prevented, which helps to alleviate the discomfort given to the subscribers.

[Brief description of the drawings]

1 is a block diagram showing the principle of the present invention, FIG. 2 is a diagram showing an embodiment of the present invention, FIG. 3 is a transmission image control circuit diagram, FIG. FIG. 2 is an explanatory diagram of a conventional videophone conference device. 1 to 3, i ₁ is an ISDN audio channel, i ₂ is an ISDN image channel, 2 _i is an audio CODEC, 4 is an arithmetic unit (audio power average value calculation circuit 4).
0 _i ), 6 is a decision unit (comparator 42 _i ), 8 is a decision unit (8-bit counter 41, 10-bit counter 43, counter 44 _i , 1
Bit left shift register 46 _i , full adder 48, comparator 50 _i ,
OR circuit 45, AND circuit 47, registers 50) and 10 are output image code registers.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Yoshihisa Mizuno 1015 Uedanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture Inside Fujitsu Limited (72) Inventor Shinichi Maki 1015 Uedanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture Fujitsu Limited (56) References JP-A-56-103566 (JP, A) JP-A-60-247362 (JP, A) JP-B-45-21203 (JP, B1)

Claims (1)

(57) [Claims] 1. A videophone conference apparatus to which each videophone terminal is connected using one of two communication channels provided by ISDN as a voice communication channel and the other communication channel as a video communication channel. For each ISDN channel among the plurality of ISDN channels, a plurality of voice CODECs connected to a channel (hereinafter abbreviated as an ISDN voice channel) allocated to transmission of a voice signal in the ISDN channel; It is connected to a channel (hereinafter abbreviated as an ISDN image channel) allocated to the transmission of an image signal, selects a desired image from the video conference call participants images, and sets the selected image to the ISDN image channels of all the participants. Output image code selector to be output and connected to the output of each audio CODEC, and output the average audio power value for each average power judgment period for each participating subscriber And a determining unit for determining whether or not each audio power average value exceeds a predetermined threshold value, and responding to the determination output corresponding to each participant output from the determining unit. For each utterance probability calculation period that is a predetermined multiple of the average power determination period for the determination, a determining unit that determines one participating subscriber whose utterance probability is equal to or greater than a predetermined value is provided. A video-telephone conference apparatus for causing a selection operation of the output image code selector.