JPH0983412A - Digital echo canceller - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the precision of the echo canceller of a type not always updating a filter coefficient by recognizing a change in a position of a microphone and the movement of persons in a room or the like so as to update the filter coefficient as required. SOLUTION: A microphone 1 is a movable desk top microphone, a sensor 7 is provided to a bottom of the microphone and its state changes when the microphone 1 is placed on a desk or the microphone 1 is apart from the desk. When a system controller 5 senses a change in the state of the sensor 7, it is recognized that the microphone 1 is moved, that is, the acoustic characteristic is changed and an instruction is given to an echo canceller 6 to update the filter coefficient.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital echo canceller device, and more particularly, to a digital echo canceller device, which is suitable for canceling echoes and howlings caused by acoustic coupling between a microphone and a speaker. The present invention relates to an echo canceller device.

[0002]

2. Description of the Related Art In satellite communication or an international telephone line network using a transoceanic submarine cable, an echo is generated due to a propagation delay of a voice signal, which makes it difficult to communicate. As the device for controlling the echo, an echo suppressor (echo suppressor) that inserts a loss in the echo path by a voice switch to forcibly stop the echo, and an echo that adaptively estimates the echo signal and subtracts it from the echo There is a canceller (echo canceller), and the echo canceller has a pseudo echo generation unit equivalent to the path in which the echo wraps around, creates a pseudo echo through the received voice signal, and converts this pseudo echo from the actual echo. It is to subtract and cancel the echo.

Most of the processing of the echo canceller concentrates on the update of the filter coefficient for generating the pseudo echo by the filter processing. From the processing amount, the echo canceller has high performance such as DSP (digital signal processing). This requires hardware, which increases the cost of the device.

Further, in a voice or video conference system, a communication device equipped with an echo canceller is used. However, in the conventional echo canceller, the amount of echo is changed as the position of the microphone changes or a person moves indoors. When the sound field changes suddenly, the updating process of the filter coefficient cannot follow, the echo component cannot be canceled, and many residual echoes are generated, which causes howling.

As an invention to solve such a problem, there is an invention disclosed in Japanese Patent Laid-Open No. 6-326761. According to the present invention, when the echo or sound field in the conference room changes abruptly, the updating process of the filter coefficient in the pseudo echo generation unit cannot follow the change, and the residual echo data output from the synthesis unit is When it grows temporarily,
The residual echo detection unit monitors the residual echo with a certain threshold value, notifies the host CPU when the threshold value is exceeded, and this host CPU controls the volume for adjusting the volume of the speaker and lowers the volume of the speaker or the host CPU. CPU
By reducing the volume of the speaker by a method of inputting the value of digital data input to the D / A converter in proportion to the original value, howling caused by a sudden change in the sound field can be prevented. It was done like this.

[0006]

Since the acoustic characteristics using the echo canceller device should not change largely in principle, the filter coefficient is not always updated, and a certain trigger is generated when the acoustic characteristics change. If the filter coefficient is updated only at, the accuracy will be slightly worse, but since it can be established as an echo canceller,
The processing amount of the echo canceller as a whole is drastically reduced.

However, the conventional echo canceller device is
There is nothing but updating the filter coefficient only at the start of voice communication of a video conference device (as so-called training), and there is no method using the change of the actual acoustic characteristic as a trigger for the update of the filter coefficient.

Further, according to the invention described in the above-mentioned Japanese Patent Laid-Open No. 6-326761, when the update processing of the filter coefficient becomes necessary, the reproduction volume is lowered when the update of the filter coefficient cannot follow the change of the acoustic characteristic. By doing so, howling and the like are prevented. Thus, the conventional echo canceller device does not positively set the trigger condition for updating the filter coefficient.

The present invention has been made in view of the above situation, and recognizes a change in the position of a microphone or a movement of a person in a room so that the filter coefficient can be updated when necessary. Thus, the object is to improve the accuracy of the echo canceller device of the type that does not constantly update the filter coefficient.

[0010]

According to a first aspect of the present invention, a movable tabletop microphone for inputting voice and a microphone arranged at the bottom of the microphone are detected. And a sensor for detecting movement of the microphone when the microphone recognizes that the microphone has moved away from the tabletop, and updates the parameters necessary for the echo canceller operation. Is provided, the movement of the microphone is detected, and this is used as a trigger for the change of the sound field, so that the filter coefficient can be updated at an optimum timing.

According to a second aspect of the present invention, there is provided a voice level detecting means for detecting the level of the voice inputted from the microphone, and when the voice level detected by the voice level detecting means exceeds a certain threshold value, It is characterized by updating the parameters required for echo canceller operation,
The input voice level that is considered to have moved the microphone is set as a threshold (threshold), the input voice level is monitored, and the threshold is used as a trigger for a change in the sound field, so that the filter coefficient is set at the optimum timing. It is made possible to update.

According to a third aspect of the present invention, in a digital echo canceller device used in a video conference system or the like, there is provided image data analysis means for analyzing a change in image data to be transmitted, and the analysis result of the image data analysis means. On the basis of the above, when it is recognized that an object has moved in the image, the parameter necessary for the echo canceller operation is updated, the content of the image of the video conference is analyzed, and the motion of the image is analyzed. Is used as a trigger for the change of the sound field, so that the filter coefficient can be updated at an optimum timing.

[0013]

1 is a block diagram for explaining an embodiment of a digital echo canceller device according to the present invention, in which 1 is a microphone, 2 is a speaker, 3 is an A / D converter, and 4 is Is a D / A converter, 5 is a system controller, 6 is an echo canceller, and 7 is a sensor.

The microphone 1 is a movable tabletop microphone, the sensor 7 is provided at the bottom of the microphone, and the state changes when the microphone 1 is placed on the table and when it is separated from the table. .

The voice data is taken in from the microphone 1, converted into digital data by the A / D converter 3, and then transferred to the echo canceller 6 for EC (echo canceller) processing, and further system control. Device 5
Will be transferred to and used accordingly.

The filter coefficient used in the echo canceller 6 is updated at the start of operation of the echo canceller device, such as when the power of the echo canceller device is turned on. However, when an object moves in a room where the echo canceller device is used,
Since the acoustic characteristics related to the echo canceller hardly change unless the number is increased or decreased or the positions of the microphone and the speaker are changed, once the filter coefficient is set, the filter coefficient is not updated.

In the example shown in FIG. 1, when the system controller 5 detects a change in the sensor 7, it recognizes that the microphone 1 has moved, that is, changes the acoustic characteristics, and issues an instruction to the echo canceller 6. The filter coefficient is updated.

FIG. 2 is a diagram for explaining another embodiment of the digital echo canceller device according to the present invention. In FIG. 2, 16 is an echo canceller, and other parts having the same operation as in FIG. The same reference numerals are given.

The example shown in FIG. 2 is an echo canceller 16.
, The input voice level from the microphone 1 is monitored, and when a certain threshold (threshold value) is exceeded, that is, when the microphone 1 moves, it is recognized that the acoustic characteristics have changed, and the filter coefficient Is to be updated.

FIG. 3 is a diagram for explaining still another embodiment of the digital echo canceller device according to the present invention.
In the example used in a video conference device that requires image processing, in the figure, 8 is a camera, 9 is an image change recognition device, and other parts having the same operations as those in FIGS. The same reference numeral as 2 is attached.

In the example shown in FIG. 3, the image captured by the camera 8 is the sound field that is the target of the echo canceller 6, and the image data is subjected to digital codec processing.
It is recognized whether or not each point on the screen has changed, and using this image data, the image change recognition device 9 determines whether or not there is a movement across the screen at the same time as the image codec, and informs the system control device 5 of the image. If it is determined that there is movement across the room, it is recognized that a person has moved inside the room, that is, the acoustic characteristics have changed, and the system control device 5
Is for instructing the echo canceller 6 to update the filter coefficient.

[0022]

According to the invention of claim 1, a movable tabletop microphone for inputting voice and a sensor arranged at the bottom of the microphone for detecting that the microphone is placed on the tabletop. When the sensor recognizes that the microphone is separated from the tabletop, the parameters necessary for the echo canceller operation are updated. Is detected,
Accordingly, it is possible to inexpensively realize an echo canceller capable of recognizing a change in acoustic characteristics and updating a filter coefficient at an optimum timing.

According to a second aspect of the present invention, there is provided voice level detection means for detecting the level of voice input from the microphone, and when the voice level detected by the voice level detection means exceeds a certain threshold value, It is characterized by updating the parameters required for echo canceller operation,
The voice level generated when the microphone moves is used as a threshold value to detect the input voice level and detect the movement of the microphone. By this, changes in acoustic characteristics are recognized and the filter coefficient is updated. An echo canceller that can be performed at an optimum timing can be realized without using a special additional device.

According to a third aspect of the present invention, in a digital echo canceller device used in a video conference system or the like, there is provided image data analysis means for analyzing a change in image data to be transmitted, and the analysis result of the image data analysis means. Based on the above, when it is recognized that an object has moved in the image, the parameter necessary for the echo canceller operation is updated, and the position of the person in the image which is the sound field is moved. By recognizing, the change in the acoustic characteristics is recognized, whereby the filter coefficient of the echo canceller can be updated at the optimum timing.

[Brief description of drawings]

FIG. 1 is a diagram for explaining an embodiment of a digital echo canceller device according to the present invention.

FIG. 2 is a diagram for explaining another embodiment of the digital echo canceller device according to the present invention.

FIG. 3 is a diagram for explaining still another embodiment of the digital echo canceller device according to the present invention.

[Explanation of symbols]

1 ... Microphone, 2 ... Speaker, 3 ... A / D converter, 4
... D / A converter, 5 ... System controller, 6, 16
... echo canceller, 7 ... sensor, 8 ... camera, 9 ... image change recognition device.

Claims (3)

[Claims] 1. A movable tabletop microphone for inputting voice, and a sensor arranged at the bottom of the microphone for detecting that the microphone is placed on a tabletop. When it recognizes that the microphone has moved away from the tabletop,
A digital echo canceller device for updating parameters required for echo canceller operation. 2. A voice level detecting means for detecting the level of voice input from a microphone, which is required for an echo canceller operation when the voice level detected by the voice level detecting means exceeds a certain threshold value. Echo canceller device characterized by updating various parameters. 3. A digital echo canceller device used in a video conference system or the like, comprising image data analysis means for analyzing changes in image data to be transmitted, and based on an analysis result of the image data analysis means. A digital echo canceller device characterized by updating a parameter required for an echo canceller operation when it is recognized that an object moves in an image.