JPH0698319A - Voice tracing type camera focusing system - Google Patents

Abstract

(57) [Summary] [Purpose] With regard to the audio tracking type camera aiming method in a video conference system, the purpose is to be able to faithfully display the status of discussions by multiple attendees in another room. [Structure] A voice detection unit 200 for detecting a speech of each attendant 100, a voice source position information storage unit 300 for storing position information of each voice source detected by the voice detection unit, and a voice detected by the voice detection unit. (By giving a stop delay of a predetermined time, etc.), the voice source identification means 400 for identifying one or a plurality of simultaneous speakers, the identification result of the voice source identification means, and the voice source position information storage means have already been stored. Based on the position information, aiming information creating means 5 for creating aiming information necessary for aiming by the TV camera 600 installed at a predetermined position including all speakers and transmitting the aiming information to the TV camera.
00 and 00 are provided.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice tracking type camera aiming system in a video conference system. 2. Description of the Related Art In recent years, a video conference system has been put into practical use, in which audio and video of each conference held in a plurality of conference rooms are mutually transferred and displayed, and an effect equivalent to that of a conference in which all members participate is exhibited.

In this type of video conference system,
It is desirable that the discussion status by multiple speakers in each conference room be displayed in the other room as faithfully as possible.

[0003]

2. Description of the Related Art In the original video conference system,
In order to capture the debate situation in the conference room as faithfully as possible, it was considered to operate the TV camera by a dedicated operator, but in recent years, a voice tracking type video conference system has been put to practical use for the purpose of eliminating the operator. Has been done.

In a conventional voice-tracking type video conference system, the utterance of a participant in a conference is detected by a microphone or the like, and one attendee who detects the utterance is aimed at the camera.
The video of the speaker was transferred to another room and displayed. [In the text, controlling a camera to capture an image of a required range including a predetermined target is referred to as "aiming"].

Therefore, when a plurality of attendees speak, the volume of each detected speech must be compared and one attendee with the highest volume must be aimed at.

[0006]

As is clear from the above description, in the conventional voice-tracking type video conference system, even when a plurality of attendees speak at the same time for discussion, the volume of the volume is the highest. Since the video of one large attendee was transferred and displayed in another room, there was a problem that the situation of discussions by multiple attendees could not be faithfully displayed in another room.

It is an object of the present invention to faithfully display the status of discussions by a plurality of attendees in another room.

[0008]

FIG. 1 is a diagram showing the principle of the present invention. In FIG. 1, 100 is a participant in the conference,
Reference numeral 600 denotes a TV camera installed at a predetermined position in the conference room.

Reference numeral 200 is a voice detecting means provided by the present invention. Reference numeral 300 is a voice source position information storage means provided by the present invention. Reference numeral 400 is a vocal source identifying means provided by the present invention. Reference numeral 500 is aiming information creating means provided by the present invention.

[0010]

The voice detecting means 200 detects the speech of each attendee 100. The utterance source position information storage unit 300 stores the position information of each utterance source detected by the voice detection unit 200.

The voicing source identifying means 400 is a voice detecting means 2
The utterances detected by 00 identify one or more simultaneous speakers. The aiming information creating means 500 includes the identification result of the utterance source identifying means 400 and the utterance source position information storing means 3.
Based on the position information stored in 00, aiming information necessary for aiming by the TV camera 600 including all speakers is created and transmitted to the TV camera 600.

It should be noted that the voicing source identifying means 400 uses the attendees 100 until a predetermined period elapses after the speech of any of the attendees 100 detected by the voice detecting means 200 is stopped.
Considering as a speaker is considered.

Therefore, the discussion status in which a plurality of attendees are speaking can be displayed in another room, and the convenience of the video conference system is greatly improved.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. 2 is a diagram showing a video conference system according to an embodiment of the present invention, FIG. 3 is a diagram explaining aiming information according to an embodiment of the present invention, and FIG. 4 is a diagram showing an example of an image in FIG. Is. The same reference numerals denote the same objects throughout the drawings.

2 to 4, the attendee 2 is shown as the attendee 100 in FIG. 1, and the microphone (M) 3 as the voice detecting means 200 in FIG. 1 and the voice receiving unit 51 in the central control unit 5 are shown. 1, the basic information section 61 and the individual information section 62 are provided in the main storage device 6 as the utterance source position information storage means 300 in FIG. 1, and the utterance source detection section 52 is provided as the utterance source identification means 400 in FIG. Is provided in the central controller 5, and the focus calculation section 5 is used as the aiming information creating means 500 in FIG.
4, a sighting control unit 55 and an attitude control unit 56 are provided in the central control unit 5, and the television camera 60 in FIG.
The camera 4 and the camera drive unit 41 are provided as 0.

In FIG. 2, video conferencing systems 10 (individual video conferencing systems are referred to as 10 _{1 to} 10 _n ) respectively installed at n remote locations are interconnected by an information communication network 20, Among them, only the configuration of the video conference system 10 ₁ is displayed in detail, and the configurations of the other video conference systems 10 _{2 to} 10 _n are omitted.

[0017] in a conference room with a TV conference system 10 ₁ has been installed, in the vicinity of the conference table 1 of the "U" shape, referred to as the 2 ₁ or 2 ₉ attendees 2 (individual attendees of 9 people , And so on) will sit down and hold a conference.

A dedicated microphone (M) 3 corresponding to each attendee 2 is installed on the conference desk 1, and a camera 4 is installed in front of the conference desk 1. .

The positions of the microphones (M) 3 and the camera 4 are unchanged. As shown in FIG. 3, the position of each object in the conference room is perpendicular to the floor with the x-axis and y-axis parallel to the floor, with the camera 4 (lens) as the origin. Three-dimensional coordinates (x, y,
z).

On the other hand, in the main storage device 6, the basic information section 6
1, the individual information section 62 and the aim recording section 63 are provided,
The basic information section 61 includes the conference room and the conference to be held.
Basic information about the video conferencing system, for example
Meeting room size that represents the size of the meeting room
(X_max), (Y_max) And (z_max) And each microphone
Including attendees 2 from the installation position of Rohon (M) 3
X-axis corrected distance (Δx) and z-axis corrected distance
With the separation (Δz), the video conferencing system outputs an audio signal.
Maximum volume value (V_max) And
And minimum volume value (V_min) And are stored, and also individually
In the information section 62, each microphone (M) 3₁Through 3 ₉of
Position vector (F₁[X₁, Y₁,
z₁]) To (F₉[X ₉, Y₉, Z₉]) Is stored
ing.

The aiming recording section 63 stores aiming information previously created by the video conference system, that is, aiming information indicating the present aiming state of the camera 4. Each microphone (M) 3 receives the voice signal (V) uttered by the corresponding attendee 2 and independently transmits the voice signal (V) to the voice receiving unit 51 in the central control device 5.

The voice receiving unit 51 includes microphones (M).
The voice signals (V) transmitted from the A. 3 are independently transmitted to the voicing source detection unit 52, and the respective voice signals are combined and transmitted to the input / output control unit 53.

Attendee 2₁And 2₇Are emitted at the same time
If you say, microphone (M) 3₁And 3₇But
Both are audio signals (V₁) And (V₇) Receive it
Independently transmitted to the voice receiving unit 51, and also transmitted to the voice receiving unit 51.
Each voice signal (V₁) And (V₇) Each independently
Each voice signal is transmitted to the voicing source detector 52.
(V ₁) And (V₇) Are synthesized into the input / output controller
introduce.

The voicing source detection unit 52 uses the voice signals (V ₁ ) and (V ₇ ) transmitted from the voice receiving unit 51 as the maximum volume value (V _max ) and the minimum volume value stored in the basic information unit 61. When comparing with the value (V _min ) and recognizing that both are within the range of a normal audio signal, it is recognized that the attendees 2 ₁ and 2 ₇ simultaneously speak, and the microphones stored in the individual information section 62 are recognized. (M) 3 ₁ and 3 ₇ position vectors (F ₁ [x ₁ , y ₁ , z ₁ ]) and (F ₇ [x
₇ , y ₇ , z ₇ ]) are extracted and transmitted to the focus calculation unit 54.

First, the focus calculation unit 54 stores the attendees 2 ₁ and 2 ₇ recognized as speaking in the same image.
The focal length (L _f ) of the camera 4 is calculated. First, the focus calculation unit 54 receives the two position vectors (F ₁ [x ₁ , y ₁ , z ₁ ]) and (F ₁ ) transmitted from the vocal source detection unit 52.
₇ [x ₇ , y ₇ , z ₇ ]), each x-axis component (x ₁ ) and (x ₇ ), y-axis component (y ₁ ) and (y ₇ ), and z-axis component (z ₁ ) and The average value of each (z ₇ ) (x
_f ), (y _f ) and (z _f ) are calculated, and a position vector (x _f ), (y _f ) and (z _f ) respectively having a mean value (x _f ), a y axis component and a z axis component ( F _f
[X _f , y _f , z _f ]) defined by a point (P [x _f , y
_f , z _f ]) is defined as the focal point (P _f ).

[0026] Then the focus calculating section 54, the camera 4 of the focal length of the (coordinate origin) until the focus (P _f) (L _f), that is, the position vector (F _f [x _f, y _f, z _f]) the length, the components (x _f), calculated using the (y _f) and (z _f).

Next, the focus calculation section 54 sets the x-axis angle (θ _x ) and the z-axis indicating the aiming direction of the camera 4 so that the attendees 2 ₁ and 2 ₇ recognized as speaking can be included in the same image. The angle (θ _z ) and the x-axis range angle (Δθ _x ) and the z-axis range angle (Δθ _z ) indicating the aiming range of the camera 4 are calculated.

First, the focus calculation unit 54 extracts the x-axis correction distance (Δx) from the basic information unit 61, and uses the extracted x-axis correction distance (Δx) to detect the position vector (F) of the microphones (M) 3 ₁ and 3 _7. By correcting the x-axis components of ( ₁ ) and (F ₇ ), the position vectors (F _1DX [x ₁ −Δx, y ₁ , z ₁ ]) and (F _7DX [including attendees 2 ₁ and 2 _7] are corrected. x ₇ + Δx, y ₇ , z ₇ ]), and the obtained position vectors (F _1DX [x ₁ −Δx, y ₁ ,
z ₁ ]) and (F _7DX [x ₇ + Δx, y ₇ , z ₇ ])
X-axis angles (θ _x1 ) and (θ _x7 ) and z-axis angles (θ _z1 ) and (θ _z7 ), and two further obtained x-axis angles (θ _x1 ) and (θ _x7 ). Average value (θ _x )
And the average value (θ _z ) of the two z-axis angles (θ _z1 ) and (θ _z7 ), and the calculated average values (θ _x ) and (θ _z ) are the x-axis angles indicating the sighting direction. (Θ _x ) and z
Axial angle (θ _z ).

Subsequently, the focus calculator 54 indicates the aiming direction.
x-axis angle (θ_x) And the position vector (F_1DX) X-axis angle
Degree (θ_x1) Or (F_7DX) X-axis angle (θ_x7) Difference
Angle (θ_x−θ_x1) Or (θ_x7−θ_x) Asked for
Difference angle (θ_x−θ_x1) = (Θ_x7−θ_x) Is the x-axis range angle
Degree (Δθ_x), And the z-axis angle indicating the aiming direction
(Θ _z) And the position vector (F_1Dz) Z-axis angle
(Θ_z1) Or (F_7Dz) Z-axis angle (θ_z7) Difference angle
(Θ_z−θ_z1) Or (θ_z7−θ_z) Asked for
Difference angle (θ_z1−θ_z) = (Θ_z−θ_z7) Is the z-axis range angle
Degree (Δθ_z).

As described above, the focus calculation section 54 uses the focal length (L _f ) and the x-axis angle (L _f ), as the aiming information necessary for keeping the attendees 2 ₁ and 2 ₇ determined to be speaking in the same image. θ _x ), z-axis angle (θ _z ), x-axis range angle (Δθ
_x ) and the z-axis range angle (Δθ _z ) are calculated.

The focus calculation unit 54 calculates the calculated focal length (L
_f ), the x-axis range angle (Δθ _x ) and the z-axis range angle (Δ
θ _z ) is transmitted to the aiming control unit 55, and the x-axis angle (θ _x ) and the z-axis angle (θ _z ) are transmitted to the attitude control unit 56.

The aiming control unit 55 is stored in the aiming recording unit 63.
The aiming information that is being provided, that is, the aiming status of the camera 4
From quasi information to focal length (L_f), X-axis range angle (Δθ_x)
And z-axis range angle (Δθ_z) Is extracted and the focus calculation unit 5
Focal length (L_f), X-axis range angle (Δ
θ_x) And the z-axis range angle (Δθ_z) With
Then, after transmitting it to the camera drive unit 41,
Stored focal length (L_f), X-axis range angle (Δθ
_x) And the z-axis range angle (Δθ_z), The focus calculation unit 54
Focal length (L_f), X-axis range angle (Δθ
_x) And the z-axis range angle (Δθ_z) Updated by
The attitude control unit 56 uses the aim recording unit 63 to store the aim
X from the aiming information, that is, aiming information indicating the present aiming state of the camera 4.
Axis angle (θ _x) And the z-axis angle (θ_z) Extract and focus
The x-axis angle (θ_x) And z-axis
Angle (θ_z) Is calculated and transmitted to the camera drive unit 41.
After reaching, the x-axis angle stored in the aim recording unit 63
(Θ_x) And the z-axis angle (θ_z) From the focus calculation unit 54
X-axis angle (θ_x) And the z-axis angle (θ_z)
To update.

The camera drive unit 41 has the focal length (L _f ) and the x-axis range angle (Δθ _x ) transmitted from the aiming control unit 55.
And the z-axis range angle (Δθ _z ) are used to update the focal length (L _f ) and aiming range of the camera 4, and the x-axis angle (θ _x ) and the z-axis angle (θ) transmitted from the attitude control unit 56. The aiming direction of the camera 4 is updated based on the difference of _z ).

The camera 4 having the updated sighting condition is shown in FIG.
Attendee 2 as shown₁And 2 ₇A video that includes
Image created in the central controller 5 as a video signal (I)
It is transmitted to the receiving unit 57.

The image receiving unit 57 transmits the image signal (I) transmitted from the camera 4 to the input / output control unit 53.
The input / output control unit 53 combines the audio signals (V ₁ ) and (V ₇ ) transmitted from the audio receiving unit 51 with the video signal (I) transmitted from the video receiving unit 57, and connects the information communication network 20.
And is transmitted to other video conference systems 10 _{2 to} 10 _n via.

The input / output control unit 53 also receives the audio signal (V) and the video signal (I) (combined) transmitted from the other video conference systems 10 _{2 to} 10 _n via the information communication network 20. Then, the audio signal (V) and the video signal (I) are separated, and the audio signal (V) is output to the speaker 31 so that each attendee 2 in the conference room can hear the video signal (I).
Is output to the monitor 42 and displayed to each attendee 2.

As is apparent from the above description, according to the present embodiment, the video conference system 10 ₁ detects the speakers 2 ₁ and 2 ₇ and calculates aiming information such that an image including both the speakers is photographed. Since the camera 4 is aimed at, the discussion status of the speakers 2 ₁ and 2 ₇ is faithfully transferred and displayed on one screen to the other video conference systems 10 _{2 to} 10 _n .

It is to be noted that FIG. 2 is merely an embodiment of the present invention, and for example, the voicing source detection unit 52 is limited to the one for identifying the attendee 2 who is speaking only by the presence or absence of the voice signal (V). However, after the voice signal of each attendee 2 is stopped, it is treated as a speaker for a predetermined elapsed time, so that all the speakers can be treated while a plurality of attendees 2 alternately speak. Many other modifications are considered, such as holding the included image, but the effect of the present invention does not change in any case.
It goes without saying that the video conference system to which the present invention is applied is not limited to the one shown in the figure.

[0039]

As described above, according to the present invention, in the video conference system, the discussion status in which a plurality of attendees are speaking can be displayed in another room, and the convenience of the video conference system is greatly improved. To do.

[Brief description of drawings]

FIG. 1 is a diagram showing the principle of the present invention.

FIG. 2 is a diagram showing a video conference system according to an embodiment of the present invention.

FIG. 3 is a diagram illustrating aiming information according to an embodiment of the present invention.

FIG. 4 is a diagram showing an example of an image in FIG.

[Explanation of symbols]

 1 Conference Desk 2, 100 Attendees 3 Microphone (M) 4 Camera 5 Central Control Device 6 Main Storage Device 10 Video Conference System 20 Information Communication Network 31 Speaker 41 Camera Drive Unit 42 Monitor 51 Voice Receiver 52 Voice Source Detector 53 In Output control unit 54 Focus calculation unit 55 Aiming control unit 56 Posture control unit 57 Image receiving unit 61 Basic information unit 62 Individual information unit 63 Aiming recording unit 200 Voice detecting means 300 Vocal source position information storing means 400 Vocal source identifying means 500 Aiming information Creating means 600 TV camera

Claims (2)

[Claims] 1. In a video conference system for transferring and displaying an image of a conference room to another room, a voice detecting means (2) for detecting a speech of each attendant (100).
00), a voice source position information storage unit (300) that stores the position information of each voice source detected by the voice detection unit (200), and a speech detected by the voice detection unit (200). Based on a voice source identification unit (400) for identifying a plurality of simultaneous speakers, an identification result of the voice source identification unit (400), and the position information stored in the voice source position information storage unit (300). Aiming information creating means (500) for creating aiming information necessary for aiming by the TV camera (600) installed at a predetermined position including all speakers and transmitting the aiming information to the TV camera (600). A voice tracking type camera aiming system characterized by providing and. 2. The voicing source identifying means (400) is any attendee (10) detected by the voice detecting means (200).
The voice tracking camera aiming system according to claim 1, wherein the attendant (100) is regarded as a speaker until a predetermined period of time elapses after the speech of (0) is stopped.