JP2006333062A - Howling prevention method, device therefor, program therefor, and recording medium therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To realize a howling prevention without degrading a two-way communication performance even for a received signal having a band in which an audio signal such as a voice signal and a musical signal is not included. <P>SOLUTION: An input signal and the received signal are each divided into a plurality of frequency bands to find attenuation quantities for the respective frequency bands. The respective attenuation quantities are distributed to attenuating means of a howling prevention device which attenuate the input signal and received signal having been filtered by the frequency bands, and signals obtained by the respective attenuating means are added together to obtain a transmitted signal and an output signal. Further, an attenuation quantity which is found for each frequency band and causes no howling may be an attenuation quantity which is smaller for a band of lower frequency. <P>COPYRIGHT: (C)2007,JPO&amp;INPIT

Description

  The present invention relates to a howling prevention method and apparatus for hands-free communication such as a TV conference and an audio conference.

A conventional howling prevention apparatus and method will be described.
FIG. 19 is a functional block diagram of a howling prevention device in the prior art disclosed in Patent Document 1. In FIG. A conventional howling prevention apparatus is configured by an acoustic coupling amount estimation unit 901, a transmission / reception determination unit 902, an attenuation control unit 903, a reception-side attenuator 906R, and a transmission-side attenuator 906S. .

  In the conventional howling prevention apparatus shown in FIG. 19, for convenience of explanation, two-way voice communication by speaker A and speaker B is illustrated. The “microphone sound reception signal” An input signal obtained by receiving a sound by a microphone, which is a sound receiving device, is represented. “Transmission signal” refers to a transmission in which a microphone sound reception signal is subjected to processing described later and is sent to the speaker B side. It shall represent a signal. Further, the “received signal” represents a transmitted signal transmitted from the speaker B side, and the “speaker output signal” represents an output signal input to a speaker serving as an output device. To do.

The acoustic coupling amount estimation unit 901 estimates the acoustic coupling amount from the speaker output signal and the microphone sound reception signal. The acoustic coupling amount is an amplitude value of a transfer function between the speaker and the microphone. When the frequency domain transformation of the speaker output signal is X (ω) and the frequency domain transformation of the microphone sound reception signal is Y (ω), the acoustic coupling amount AC (ω) can be obtained by Expression (1). However, | · | represents an absolute value.

  The transmission / reception determination unit 902 determines transmission / reception from the reception signal and the microphone reception signal. In the reception detection, the level of the reception signal is observed, and when the level exceeds a fixed threshold given in advance and a threshold obtained by multiplying the noise level by a constant, it is determined as reception. In the transmission detection, the level of the microphone sound reception signal is observed, the level exceeds a fixed threshold value obtained in advance and a threshold value obtained by multiplying the noise level by a constant, and the absolute value of the speaker output signal is added to the acoustic coupling amount AC (ω). If the estimated acoustic echo level multiplied by | x (ω) |

  In a silent section in which neither transmission nor reception is detected, or in a double talk section in which both transmission and reception are detected, determination is made as either transmission or reception from the previous transmission / reception determination state. In order to prevent frequent switching of transmission / reception determination, a hangover time is provided so that the transmission determination state or the reception determination state is maintained for a certain period of time.

The attenuation amount control unit 903 obtains an attenuation amount that does not cause howling from the acoustic coupling amount AC (ω), and distributes the attenuation amount to the transmission side gain gs and the reception side gain gr from the transmission / reception determination result. First, a sufficient condition for preventing howling is that the amplitude of the transfer function from reception to transmission is less than 1 for all frequencies. If this condition is expressed by an equation, equation (2) is obtained. However, MAX _ω {·} represents that the maximum value for ω is taken.

Therefore, the amount of attenuation for preventing howling is expressed by equation (3). However, C is a constant less than 1 and is given in advance.

Next, the distribution to the gain gs on the transmission side and the gain gr on the reception side is determined from the transmission / reception determination result. When the transmission / reception determination unit 902 determines transmission, the transmission-side gain gs = 1 and reception-side gain gr = C / MAX _ω {AC (ω)} are determined as reception. , The transmission side gain gs = C / MAX _ω {AC (ω)}, and the reception side gain gr = 1. As a result, in the case of transmission, the transmission signal can be passed as it is, in the case of reception, the reception signal can be passed as it is, and in the case of single talk, smooth conversation is possible. In addition, since discontinuous changes in gain cause deterioration in sound quality, the gain is changed smoothly with time.
Japanese Patent No. 3268572

The conventional howling prevention apparatus and method have the following problems.
For example, in the case of IP communication or the like, the frequency band of voice to be communicated is various (up to 3.4 kHz, up to 7 kHz, up to 14 kHz, up to 20 kHz, etc.). Even if the voice codec that encodes the transmission signal and decodes the reception signal has a wide band, the frequency band may be limited by a terminal on the other side or a gateway in the communication network. At this time, in the band in which the audio signal is no longer included due to the band limitation, the estimated value of the acoustic coupling amount becomes a value larger than an accurate value.

  In such a case, in the conventional howling prevention apparatus and method, the attenuation amount is given to all the frequency bands at the same time. Therefore, the attenuation amount according to the high-frequency acoustic coupling amount that is not accurately estimated is given. End up. As a result, an unnecessarily large attenuation is given, and the two-way call performance is significantly lowered.

  In the conventional howling prevention apparatus and method, when the voice band included in the received signal is narrower than the band of the voice codec being used, the high-frequency acoustic coupling amount estimation that does not include the voice signal can be accurately performed. Can not. For this reason, an unnecessarily large attenuation amount is given to all the bands, and the two-way communication performance is deteriorated.

  SUMMARY OF THE INVENTION The problem to be solved by the present invention is to provide a howling prevention method and apparatus that does not deteriorate the bidirectional communication performance even for a received signal having a band that does not include sound signals such as voice and musical sound. .

  In order to solve the above-described problem, in the present invention, an input signal and a received signal are each divided into two frequency bands to obtain an attenuation amount for each frequency band, and each attenuation amount is filtered for each frequency band. The transmission signal and the output signal are respectively obtained by distributing the signal and the reception signal to the attenuation unit of the howling prevention device that attenuates the signal and the received signal, and adding the signals obtained by the attenuation units. In the present invention, even if a large attenuation is given to a frequency band that does not include an acoustic signal, the fact that there is no acoustic signal is used because there is no acoustic signal. The amount of acoustic coupling and transmission / reception determination are performed in the same manner as in the prior art, and the amount of attenuation to prevent howling is low (for example, up to 3.4 kz for a telephone band) and high (for example, 3 for the low band). .4kHz or more).

  In this way, even if the high-frequency acoustic coupling level cannot be accurately estimated because no acoustic signal is included in the high-frequency range, this effect is given only to the high-frequency attenuation level. The attenuation of the region can be set to an appropriate value. This makes it possible to perform communication without degrading the bidirectional communication performance.

  In the present invention, the input signal and the received signal are each divided into three or more frequency bands to obtain the attenuation amount for each frequency band. The signals are distributed to the attenuation means of the howling prevention device for attenuation, and the signals obtained by the respective attenuation means are added to obtain the transmission signal and the output signal, respectively. In this case, estimation of acoustic coupling amount and transmission / reception detection are performed in the same manner as in the prior art, and an attenuation amount for preventing howling is obtained for each divided frequency band.

  In this way, even if the acoustic coupling amount in the band that does not include the acoustic signal cannot be accurately estimated, this influence is given only to the attenuation amount in that band, and other attenuation amounts are appropriate values. It can be. This makes it possible to perform communication without degrading the bidirectional communication performance.

  Alternatively, according to the present invention, the input signal and the received signal are divided into a plurality of frequency bands, and the attenuation amount is obtained for each frequency band. The attenuation amount is attenuated for each input signal and reception signal filtered for each frequency band. Even if the processing for allocating to the attenuation means of the howling prevention apparatus to be added and obtaining the transmission signal and the output signal by adding the signals obtained by the respective attenuation means is realized by one filtering means for each of the input signal and the reception signal Good. The frequency characteristic obtained by dividing the frequency band and the attenuation amount given to each frequency band can be realized by one filtering means having the same frequency characteristic.

  By using one filter means in this way, it is possible to reduce the amount of calculation and obtain the same performance as compared to using a multi-channel band division filter.

  Furthermore, in the present invention, an attenuation amount that does not cause howling that is required for each frequency band may be set to a smaller attenuation amount in a lower frequency band. When the attenuation amount in the lower band is large and the attenuation amount in the higher band is small, only the signal in the high band is communicated, and there is a possibility that the quality of the signal is deteriorated. In order to prevent this, it is possible to prevent the signal quality from deteriorating by making the attenuation amount in the higher band larger than the attenuation amount in the band lower than the frequency band.

  Further, from another point of the present invention, frequency bands are divided using short-time Fourier transform, attenuation amounts are obtained for the respective frequency bands, and the respective attenuation amounts are input to the input signal filtered for each frequency band and The received signal may be distributed to the attenuating means of the howling prevention device for attenuating each signal, and the transmission signal and the output signal may be obtained by performing inverse Fourier transform on the signal obtained by each attenuating means for a short time. For example, when it is desired to increase the number of frequency band divisions, the frequency band division using the short-time Fourier transform can reduce the amount of calculation.

  Also, the computer can be operated as a howling prevention device by a howling prevention program that causes the howling prevention device of the present invention to function on the computer. Then, the computer-readable program recording medium that records this howling prevention program enables other computers to function as a howling prevention device, distribute the howling prevention program, and the like.

  According to the present invention, an attenuation amount for preventing howling is given to each of an input signal and a received signal for each frequency band. It is possible to prevent howling without degrading performance.

  In each of the following embodiments, for convenience of explanation, a case where the howling prevention apparatus / method of the present invention is used for two-way voice communication by a speaker C and a speaker D will be described as an example. Accordingly, the “microphone sound reception signal” in the following description and drawings represents an input signal obtained by receiving a sound such as a speech of the speaker C by a microphone which is a sound reception device. The “signal” represents a transmission signal that is transmitted to the speaker D side after the microphone reception signal undergoes the processing described later. The “reception signal” is a transmission signal that is transmitted from the speaker D side to the speaker C side. The “speaker output signal” represents an output signal input to a speaker which is an output device. Each signal is a digital signal that has been appropriately subjected to known sampling processing and quantization processing, and is necessary for executing A / D conversion to an input signal, D / A conversion to an output signal, and the like. The constituent elements (means) are all omitted because they are achieved by conventional means in the prior art.

  Note that the two-way voice communication by two speakers (speaker C and speaker D) is not limited to two-way voice communication. Directional voice communication may be used. The sound received by the microphone is not limited to voice such as human speech, but includes all sounds such as music and noise.

  In each of the following embodiments, as an example of a howling prevention apparatus of the present invention, it will be described as being realized by a computer. Of course, the howling prevention apparatus of the present invention is not limited as being realized by a so-called general-purpose computer. For example, a DSP (Digital Signal Processor) or CPU (Central Processing Unit) that can be a processing subject in the present invention, a storage device such as a memory storing a program for executing the processing in the present invention, a handset equipped with a microphone, a speaker, and the like It can be realized as a free call device and is also preferable. In addition, since the hands-free call device in which the howling prevention method of the present invention is implemented can be realized by a configuration substantially the same as that performed by a computer described in each embodiment described later, the embodiment as a hands-free call device This will become clear from the description of each embodiment described later.

<First Embodiment>
First, a first embodiment of a howling prevention apparatus / method of the present invention will be described.
FIG. 1 is a configuration block diagram illustrating a hardware configuration of a howling prevention apparatus (1) according to the first embodiment.
FIG. 2 is a functional block diagram illustrating the processing function for preventing howling in the howling preventing apparatus 1 according to the first embodiment.
FIG. 3 is a flowchart showing a part of howling prevention processing (setting of attenuation) in the howling prevention apparatus (1) according to the first embodiment.
FIG. 4 is a flowchart showing a part of howling prevention processing (output of a transmission signal) in the howling prevention device (1) according to the first embodiment.
FIG. 5 is a flowchart showing a part of howling prevention processing (output of a speaker output signal) in the howling prevention device (1) according to the first embodiment.

  The outline of the first embodiment is that each of the microphone reception signal and the reception signal is divided into two frequency bands at a predetermined cutoff frequency (a low frequency band which is a frequency band below the cutoff frequency and a frequency band above the cutoff frequency). Attenuators that respectively attenuate the microphone sound reception signal and the reception signal filtered for each frequency band, respectively, by obtaining attenuation amounts for the low frequency band and the high frequency band. And a signal obtained by adjusting the attenuation by each attenuator is added to obtain a transmission signal and a speaker output signal. By doing so, even if the audio signal is not included in the high frequency band and the acoustic coupling amount in the high frequency band cannot be accurately estimated, the attenuation amount in the low frequency band is not affected by the high frequency band, Appropriate attenuation is obtained. Moreover, even if the attenuation amount in the high frequency band is inappropriate, the band is originally limited, so that the two-way call performance does not deteriorate and howling does not occur.

  As illustrated in FIG. 1, the howling prevention apparatus (1) according to the first embodiment includes an input unit (11) to which a keyboard or the like can be connected, and an output unit (12) to which a liquid crystal display or the like can be connected. In the case of the above-mentioned hands-free communication device, it is not always necessary. ], Howling prevention device (1) communication unit (13) to which a communication device (for example, modem) capable of communication to the outside can be connected, DSP (14) [CPU may be used. A cache memory or the like may be provided. ] RAM (Random Access Memory) (15), ROM (Read Only Memory) (16), external storage device (17) which is a hard disk, input unit (11), output unit (12), communication A bus (18) connected so that data can be exchanged between the unit (13), the DSP (14), the RAM (15), the ROM (16), and the external storage device (17). If necessary, the howling prevention device (1) may be provided with a device (drive) capable of reading and writing a storage medium such as a CD-ROM. [The read / write device for the external recording medium is the above-mentioned hands-free communication device. In some cases, it is usually unnecessary. ].

  Further, the howling prevention device (1) can be connected to a sound receiving device (for example, a microphone) that receives sound such as voice, music, and noise, and receives a signal input from the microphone. And an output device (for example, a speaker) that converts the output signal into sound and outputs it, and is provided with a signal output unit for outputting the output signal to be input to the speaker [of the above-mentioned hands-free communication device] In some cases, microphones and speakers are usually required. ]. A microphone is connected to the signal input unit, and a speaker is connected to the signal output unit.

  The external storage device (17) of the howling prevention device (1) stores and stores a program for preventing howling and data necessary for the processing of this program [of a hands-free communication device without an external storage device] In this case, for example, the program may be stored in a ROM that is a read-only storage device. ]. Further, data obtained by the processing of these programs is appropriately stored and stored in a RAM or the like.

  More specifically, in the external storage device (17), a program for realizing a filter (low-pass filter 104S, high-pass filter 105S) for filtering the microphone sound reception signal obtained by the microphone in each predetermined frequency band, A program for realizing an attenuator (106S1, 106S2) for adjusting attenuation of each filtered microphone reception signal, and an adder (107S) for adding each microphone reception signal subjected to attenuation adjustment A program for realizing a filter (low-pass filter 104R, high-pass filter 105R) for filtering a received signal received by a program, a modem or the like in each predetermined frequency band, an attenuator (106R1, for adjusting attenuation of the filtered received signal) 106R2) Program, a program for realizing an adder (107R) for adding each received signal subjected to attenuation adjustment, and an acoustic coupling amount for obtaining an acoustic coupling amount between the microphone reception signal and the output signal input to the speaker An estimation program, an attenuation control program for obtaining an attenuation amount from an acoustic coupling amount and allocating it to an attenuator, and a transmission / reception determination program for determining transmission / reception from the reception signal and the microphone reception signal are stored and stored. . In addition, a control program for controlling processing based on these programs is also stored as appropriate.

  In the howling prevention apparatus (1) according to the first embodiment, each program stored in the external storage device (17) [or ROM, etc.] and data necessary for processing each program are stored in the RAM (15 ) To be interpreted and executed / processed by the DSP (14). As a result, the DSP (14) realizes predetermined functions (acoustic coupling amount estimation unit, transmission / reception determination unit, attenuation control unit, low-pass filter, high-pass filter, attenuation unit, addition unit), thereby preventing howling. Realized.

Then, next, with reference to FIGS. 2-5, the flow of the howling prevention process in a howling prevention apparatus (1) is demonstrated sequentially.
The howling prevention apparatus (1) of the first embodiment includes an acoustic coupling amount estimation unit (101), a transmission / reception determination unit (102), an attenuation control unit (103), a low-pass filter (104R, 104S), a high-pass filter ( 105R, 105S), receiving side attenuators (106R1, 106R2), transmitting side attenuators (106S1, 106S2) and adders (107R, 107S).

The acoustic coupling amount estimation unit (101) obtains an acoustic coupling amount from the speaker output signal and the microphone sound reception signal. The acoustic coupling amount is an amplitude value of a transfer function between the speaker and the microphone. When the frequency domain conversion of the speaker output signal is X (ω) and the frequency domain conversion of the microphone sound reception signal is Y (ω), the acoustic coupling amount AC (ω) can be obtained by Expression (4). However, | · | represents an absolute value.

  The transmission / reception determination unit (102) determines transmission / reception from the reception signal and the microphone reception signal. In the reception detection, the level of the reception signal is observed, and when the level exceeds a fixed threshold given in advance and a threshold obtained by multiplying the noise level by a constant, it is determined as reception. In the transmission detection, the level of the microphone sound reception signal is observed, and the level exceeds a fixed threshold value obtained in advance and a threshold value obtained by multiplying the noise level by a constant. When the estimated acoustic echo level multiplied by the absolute value | X (ω) |

  In a silent section in which neither transmission nor reception is detected, or in a double talk section in which both transmission and reception are detected, determination is made as either transmission or reception from the previous transmission / reception determination state. In order to prevent frequent switching of transmission / reception determination, a hangover time is provided so that the transmission determination state or the reception determination state is maintained for a certain period of time.

  The attenuation amount control unit (103) obtains an attenuation amount that does not cause howling in each of the low frequency band and the high frequency band from the acoustic coupling amount AC (ω) obtained by the acoustic coupling amount estimation unit (101). The amount is distributed to the gain gs on the microphone sound reception signal side and the gain gr on the reception signal side from the transmission / reception determination result in the transmission / reception determination unit (102). This will be specifically described below.

First, a sufficient condition for preventing howling is that the amplitude of the transfer function from reception to transmission is less than 1 for all frequencies. If this condition is expressed by an equation, equation (5) is obtained.

Where MAX _ω {·} represents the maximum value for ω, and | lpf (ω) | and | hpf (ω) | are low-pass filters (104R) (104S) that pass signals in the low frequency band, respectively. ) And the frequency characteristics of the high-pass filters (105R) (105S) that pass signals in the high frequency band. Each of gs1 and gr1 is filtered by a low-frequency band microphone reception signal (hereinafter referred to as “microphone reception signal (low-pass))” filtered by the low-pass filter (104S) and a low-pass filter (104R). It represents the gain given to the received signal in the low frequency band (hereinafter referred to as “received signal (low-pass)”). Each of gs2 and gr2 is filtered by a high-frequency band microphone reception signal (hereinafter referred to as “microphone reception signal (high-pass))” filtered by the high-pass filter (105S) and a high-pass filter (105R). It represents the gain given to the received signal in the high frequency band (hereinafter referred to as “received signal (high-pass)”).

Here, for each of the low-pass filter (104R) (104S) having frequency characteristics of | lpf (ω) | and the high-pass filter (105R) (105S) having frequency characteristics of | hpf (ω) | (5) can be divided into two equations (6) and (7) for the low-frequency band and the high-frequency band.

However, MAX _ω1 {·} represents taking the maximum value for ω in the range of the low frequency band, and MAX _ω2 {·} represents taking the maximum value for ω in the range of the high frequency band.

Therefore, the attenuation amount for preventing howling in each of the low frequency band and the high frequency band is given by Expression (8) and Expression (9).

  However, it is assumed that C1 and C2 are constants less than 1 and are described in advance by being incorporated in the attenuation amount control program. These are set so as to satisfy the expressions (6) and (7) even if there is an error in the acoustic coupling amount estimation or the influence of the ripples of the low-pass filter and the high-pass filter.

In other words, the attenuation control unit (103) first uses the acoustic coupling amount AC (ω) obtained by the acoustic coupling amount estimation unit (101) to use the attenuation amount C1 / MAX _ω1 {AC (ω in the low frequency band. )} And attenuation C2 / MAX _ω2 {AC (ω)} in the high frequency band.

Next, the distribution of the attenuation amount C1 / MAX _ω1 {AC (ω)} to the gain gs1 and the gain gr1, and the distribution of the attenuation amount C2 / MAX _ω2 {AC (ω)} to the gain gs2 and the gain gr2, It is determined by the transmission / reception determination result in the transmission / reception determination unit (102).

When the transmission / reception determination result by the transmission / reception determination unit (102) is transmission, the attenuation amount control unit (103) converts the microphone reception signal (low-pass) and microphone reception signal (high-pass). The gains given are gs1 = 1 and gs2 = 1, and the gains given to the received signal (low-pass) and received signal (high-pass) are gr1 = C1 / MAX _ω1 {AC (ω)}, gr2 = C2 / MAX _ω2 {AC (ω)} is allocated, and these gains are set in attenuators (106S1) (106S2) (106R1) (106R2) described later.

When the transmission / reception determination result by the transmission / reception determination unit (102) is reception, the attenuation amount control unit (103) gives the microphone reception signal (low-pass) and the microphone reception signal (high-pass). Gains are given as gs1 = C1 / MAX _ω1 {AC (ω)}, gs2 = C2 / MAX _ω2 {AC (ω)}, and gains given to the reception signal (low-pass) and the reception signal (high-pass) It distributes as gr1 = 1 and gr2 = 1, and these gains are set in attenuators (106S1) (106S2) (106R1) (106R2) which will be described later.

In the above distribution of attenuation, for example, gs1 = 1, gr1 = C1 / MAX _ω1 {AC (ω)}, gs2 = 1, gr2 = C2 / MAX _ω2 {AC (ω)}. It is not limited to each numerical value. In short, a significant difference may be generated between gains given to signals in transmission and reception. For example, when it is determined that the transmission is performed, gs1 = 0.95, gs2 = 0.90, gr1 = C1 / {0.95 × MAX _ω1 {AC (ω)}}, gr2 = C2 / {0. 90 × MAX _ω2 {AC (ω)}}, which can be changed as appropriate.

Note that discontinuous changes in gain cause deterioration in sound quality, so it is preferable that the gain changes smoothly with time. This will be described using the gain gs1 as an example (the same applies to other gains). The attenuation control unit (103) sets the gain calculated based on the transmission / reception determination result as g (corresponding to gs1 in the above-mentioned symbol), and the gain set [in the attenuator (106S1)] one time ago. Is gs1 (t−1), a gain gs1 (t) to be newly set in the attenuator (106S1) is obtained by the equation (10), and this is set in the attenuator (106S1). However, a is a smoothing coefficient, and a value of 0 <a <1 is set in advance. The closer a is to 1, the more gentle the gain change, and the closer a is to 0, the steeper gain change. By appropriately determining the smoothing coefficient a in this way, the gain can be changed with a desired smoothness, and deterioration of sound quality due to a discontinuous change in gain can be prevented.

  Next, the attenuator (106S1) attenuates and adjusts the microphone sound reception signal (low-pass) by the gain gs1 set (distributed) by the attenuation amount control unit (103), and the attenuator (106S2) The microphone sound reception signal (high-pass) is attenuated and adjusted by the gain gs2 set (distributed) by the amount control unit (103).

  The attenuator (106R1) attenuates and adjusts the received signal (low-pass) by the gain gr1 set (distributed) by the attenuation amount control unit (103), and the attenuator (106R2) is the attenuation amount control unit. The received signal (high-pass) is attenuated and adjusted by the gain gr2 set (distributed) by (103).

When the set (distributed) gain is 1, the signal is passed without being attenuated (in other words, the attenuation amount Z (dB) of the attenuator is Z = 20 · log _{10 with} respect to the gain g). (1 / g).) That is, when gs1 = gs2 = 1, the microphone sound reception signal (low-pass) and the microphone sound reception signal (high-pass) are output without being attenuated. When gr1 = gr2 = 1, the received signal (low-pass) and the received signal (high-pass) are output without being attenuated. Thereby, in the case of transmission, the transmission signal can be passed as it is, in the case of reception, the reception signal can be passed as it is, and in the case of single talk, a smooth conversation is possible.

  Next, the adder (107S) adds and transmits the microphone sound reception signal (low-pass) that has passed through the attenuator (106S1) and the microphone sound reception signal (high-pass) that has passed through the attenuator (106S2). Ask for a speech signal. The transmission signal is transmitted to the other party (speaker D) via a communication device such as a modem.

  The adder (107R) adds the received signal (low-pass) that has passed through the attenuator (106R1) and the received signal (high-pass) that has passed through the attenuator (106R2), and outputs a speaker output signal. . The speaker output signal is input to the speaker (through D / A conversion or the like) and output as sound from the speaker.

  The above will be described below from the viewpoint of the process in which the microphone sound reception signal and the reception signal are converted into the transmission signal and the speaker output signal, respectively.

  The voice of the speaker C (including noise in the surrounding environment, etc.) is received by the microphone (via A / D conversion) and input to the howling prevention device as a microphone received signal (step S110). . On the other hand, the transmission signal transmitted from the speaker D side is received by a communication device such as a modem and input as a reception signal to the howling prevention device on the speaker C side (step S120).

  The acoustic coupling amount estimation unit (101) obtains the acoustic coupling amount AC (ω) from the microphone sound reception signal and the speaker output signal output from the adder (107R) (step S100). The transmission / reception determination unit (102) determines transmission / reception from the microphone sound reception signal and the reception signal (step S101).

The attenuation amount control unit (103) is configured to divide each frequency band (low frequency band, high frequency) by a predetermined cut-off frequency from the acoustic coupling amount AC (ω) obtained by the acoustic coupling amount estimation unit (101). Attenuation C1 / MAX _ω1 {AC (ω)} and C2 / MAX _ω2 {AC (ω)} in the band). Furthermore, the attenuation amount control unit (103) sets the gain gs1 and the attenuator for setting the attenuation amount C1 / MAX _ω1 {AC (ω)} to the attenuator (106S1) from the transmission / reception determination result by the transmission / reception determination unit (102). The gain gr1 set to (106R1) is distributed, the gain gs1 is set to the attenuator (106S1), and the gain gr1 is set to the attenuator (106R1). Further, the attenuation control unit (103) sets the gain gs2 for setting the attenuation C2 / MAX _ω2 {AC (ω)} in the attenuator (106S2) and the attenuation based on the transmission / reception determination result by the transmission / reception determination unit (102). The gain gr2 to be set in the attenuator (106R2) is distributed, the gain gs2 is set in the attenuator (106S2), and the gain gr2 is set in the attenuator (106R2) (step S102).

  The low-pass filter (104S) filters the microphone sound reception signal in a predetermined frequency band (step S111a), and the attenuator (106S1) set with the gain gs1 attenuates the filtered microphone sound reception signal. (Step S112a). The high-pass filter (105S) filters the microphone sound reception signal in a predetermined frequency band (step S111b), and the attenuator (106S2) set with the gain gs2 attenuates the filtered microphone sound reception signal. Adjust (step S112b).

  The signals output from the attenuators (106S1) and (106S2) are added by the adder (107S) and output as a transmission signal (step S113). The transmission signal is transmitted to the other party (speaker D) via a communication device such as a modem.

  The low-pass filter (104R) filters the received signal in a predetermined frequency band (step S121a), and then the attenuator (106R1) set with the gain gr1 attenuates the filtered received signal (step S122a). . The high-pass filter (105R) filters the received signal in a predetermined frequency band (step S121b), and the attenuator (106R2) set with the gain gr2 attenuates and adjusts the filtered received signal (step S121b). S122b).

  The signals output from the attenuators (106R1) and (106R2) are added by the adder (107R) and output as a speaker output signal (step S123). The speaker output signal is input to the speaker (through D / A conversion or the like) and output as sound from the speaker.

  Note that the processing in steps S100 to S102 is repeatedly executed at least as long as a reception signal or a microphone reception signal is input. Of course, the processing of steps S110 to S113 and steps S120 to S123 is repeated for each of the reception signal and the microphone reception signal input to the howling prevention apparatus.

  In the first embodiment, appropriate attenuation can be given to the low frequency band and the high frequency band of transmission / reception. Further, since the attenuation amount is obtained for each frequency band, it does not affect each other. Therefore, even if the amount of acoustic coupling in the high frequency band cannot be accurately estimated because the audio signal is not included in the high frequency band, this effect is given only to the attenuation amount in the high frequency band. The amount of attenuation can be set to an appropriate value. As a result, it is possible to make a call without degrading the two-way call performance.

<Second Embodiment>
Next, a second embodiment of the howling prevention apparatus / method of the present invention will be described.
FIG. 6 is a functional block diagram illustrating how processing for preventing howling in the howling preventing apparatus according to the second embodiment.
FIG. 7 is a flowchart showing a part of howling prevention processing (setting of attenuation) in the howling prevention apparatus according to the second embodiment.
FIG. 8 is a flowchart showing a part of howling prevention processing (output of a transmission signal) in the howling prevention device according to the second embodiment.
FIG. 9 is a flowchart showing a part of howling prevention processing (output of a speaker output signal) in the howling prevention device according to the second embodiment.

The outline of the second embodiment is that each of the microphone reception signal and the reception signal has a predetermined cut-off frequency of three or more frequency bands (a low frequency band that is a frequency band of the cut-off frequency ω ₁ or less, i = 1). , 2,..., M are divided into an i-th frequency band that is a frequency band between cut-off frequencies ω _{i and} ω _{i + 1} and a high-frequency band that is a frequency band greater than or equal to the cut-off frequency ω _{M + 1.} The number of band divisions is N (≧ 3), where M is the number of band-pass filters and M = N−2. ], The attenuation amount is obtained for each of the low frequency band, the i-th frequency band, and the high frequency band, and each attenuation amount is distributed to an attenuator that attenuates the microphone sound reception signal and the reception signal filtered for each frequency band. The signals obtained by the attenuation adjustment by the attenuators are added to obtain the transmission signal and the speaker output signal, respectively. By doing so, even if an audio signal is not included in a certain frequency band and the acoustic coupling amount in a higher frequency band cannot be accurately estimated, the attenuation amount in other frequency bands is not affected and is appropriate. Attenuation amount.

  The main part in which the second embodiment differs from the first embodiment is the number of frequency bands to be divided. Accordingly, the description of the configurations and functions common to the first embodiment will be omitted, and only the parts different from the first embodiment will be described (the hardware configuration example of the howling prevention apparatus according to the second embodiment is (Refer FIG. 1 which shows the hardware structural example of the howling prevention apparatus concerning 1st Embodiment.).

  In the external storage device of the howling prevention apparatus according to the second embodiment, in addition to the program stored and stored in the external storage device of the howling prevention apparatus according to the first embodiment, the microphone sound reception signal obtained by the microphone is stored. A program for realizing a filter for filtering in the i-th frequency band (bandpass filters 201S1 to 201SM), an attenuator (106Sk) for adjusting attenuation of each microphone sound reception signal filtered by the bandpass filters 201S1 to 201SM A program for realizing [k = 2, 3,..., N−1], and a filter (bandpass filter 201R1 to 201RM) for filtering the received signal received by the receiving device in the i-th frequency band. Program, bandpass filter 201R1 Also stored stored program for realizing attenuator for attenuation adjustment of each received signal filtered with (106Rk) by 201RM.

  In the howling prevention apparatus according to the second embodiment, each program stored in the external storage device (17) [or ROM, etc.] and data necessary for processing each program are read into the RAM (15) as necessary. Then, the interpretation is executed and processed by the DSP (14). As a result, the DSP (14) realizes predetermined functions (acoustic coupling amount estimation unit, transmission / reception determination unit, attenuation amount control unit, low pass filter, band pass filter, high pass filter, attenuation unit, addition unit), Howling can be prevented.

Then, next, with reference to FIGS. 6-9, the flow of the howling prevention process in the howling prevention apparatus concerning 2nd Embodiment is demonstrated sequentially.
A howling prevention apparatus according to the second embodiment includes an acoustic coupling amount estimation unit (101), a transmission / reception determination unit (102), an attenuation control unit (103), a low-pass filter (104R, 104S), and one or more ones. Band pass filters (201R1 to 201RM, 201S1 to 201SM), high pass filters (105R and 105S), receiver attenuators (106R1 to 106RN), transmitter attenuators (106S1 to 106SN), and adders (107R and 107S) ).

  Since the acoustic coupling amount estimation unit 101 and the transmission / reception detection unit 102 are the same as those in the first embodiment, description thereof will be omitted.

The attenuation amount control unit (103) obtains the attenuation amount of each frequency band in the same manner as in the first embodiment, and sets an appropriate gain in each attenuator. The difference from the first embodiment is the frequency. The number of bands.
The attenuation control unit (103) is an attenuation that does not cause howling in each of the low frequency band, the i-th frequency band, and the high frequency band from the acoustic coupling amount AC (ω) obtained by the acoustic coupling amount estimation unit (101). And the respective attenuation amounts are distributed to the gain gs on the microphone sound reception signal side and the gain gr on the reception signal side from the transmission / reception determination result in the transmission / reception determination unit (102). This will be specifically described below.

First, a sufficient condition for preventing howling in the second embodiment is that the amplitude of the transfer function from reception to transmission is less than 1 for all frequencies. If this condition is expressed by an equation, the equation (11) is obtained.

However, MAX _ω {·} represents the maximum value for ω, and | lpf (ω) | and | hpf (ω) | are low-pass filters (104R) (104S) that allow low-frequency band signals to pass through, respectively. ) And the frequency characteristics of the high-pass filters (105R) (105S) that pass signals in the high frequency band.

| Bpf _k-1 (ω) | represents the frequency characteristics of the bandpass filters (201R1) to (201RM) and (201S1) to (201SM) that pass signals in the (k-1) th frequency band [where, Note that k = 2, 3,..., N−1, and k = i + 1 and M = N−2. ].

  Each of gs1 and gr1 is filtered by a low-frequency band microphone reception signal (hereinafter referred to as “microphone reception signal (low-pass))” filtered by the low-pass filter (104S) and a low-pass filter (104R). It represents the gain given to the received signal in the low frequency band (hereinafter referred to as “received signal (low-pass)”).

  gsk and grk are respectively a microphone sound reception signal in the i-th frequency band filtered by the bandpass filters (201S1) to (201SM) (hereinafter referred to as “microphone sound reception signal (i-th band pass)”) and bandpass. It represents the gain given to the received signal in the i-th frequency band filtered by the filters (201R1) to (201RM) (hereinafter referred to as “received signal (i-th band pass)”).

  The gsN and grN were respectively filtered by a high frequency band microphone received signal filtered by the high pass filter (105S) (hereinafter referred to as “microphone received signal (high pass)”) and a high pass filter (105R). It represents the gain given to the received signal in the high frequency band (hereinafter referred to as “received signal (high-pass)”).

For each filter (104R, 104S, 105R, 105S, 201R1 to 201RM, 201S1 to 201SM), if the ripple in the passband is sufficiently small and the attenuation in the stopband is sufficiently obtained, equation (11) Can be divided into equations (12) for each frequency band. However, MAX _ωj {·} represents taking the maximum value for ω within the range of the j-th frequency band.

Therefore, an attenuation amount for preventing howling in each frequency band is given by Equation (13).

  However, Cj is a constant less than 1 and is preliminarily described in the attenuation amount control program. These are set so as to satisfy Expression (12) even if there is an error in acoustic coupling amount estimation or the influence of ripples of the low-pass filter and the high-pass filter.

In other words, the attenuation control unit (103) first uses the acoustic coupling amount AC (ω) obtained by the acoustic coupling amount estimation unit (101) to use the attenuation amount C1 / MAX _ω1 {AC (ω in the low frequency band. )}, Attenuation Ck / MAX _ωk {AC (ω)} in the i-th frequency band [where k = i + 1. ], The amount of attenuation CN / MAX _ωN {AC (ω)} in the high frequency band is obtained.

Next, the distribution of the attenuation C1 / MAX _ω1 {AC (ω)} to the gain gs1 and the gain gr1, and the distribution of the attenuation Ck / MAX _ωk {AC (ω)} to the gain gsk and the gain grk [where k = 2, 3,..., N−1], the distribution of the attenuation CN / MAX _ωN {AC (ω)} to the gain gsN and the gain grN depends on the transmission / reception determination result in the transmission / reception determination unit (102). It is determined.

That is, the attenuation control unit (103), when the transmission / reception determination result by the transmission / reception determination unit (102) is transmission, the microphone reception signal (low-pass), the microphone reception signal (i-th region) Pass) and the gain given to the microphone sound reception signal (high-pass) is gsj = 1 [j = 1, 2,..., N], and the received signal (low-pass) and the received signal (i-th pass) ) And gain given to the received signal (high-pass) are distributed as _grj = Cj / MAX _ωj {AC (ω)}, and these gains are distributed to attenuators (106S1 to 106SN) and (106R1 to 106RN) described later. Set.

When the transmission / reception determination result by the transmission / reception determination unit (102) is reception, the attenuation amount control unit (103) receives the microphone sound reception signal (low-frequency pass), the microphone sound reception signal (i-th frequency pass), and The gain given to the microphone sound reception signal (high-pass) is gsj = Cj / MAX _ωj {AC (ω)}, and the reception signal (low-pass), reception signal (i-th pass), and reception signal (high-pass) The gain given to (pass) is distributed as grj = 1, and these gains are set in attenuators (106S1 to 106SN) and (106R1 to 106RN) described later.

In the above-described attenuation distribution, for example, gsj = 1 and grj = Cj / MAX _ωj {AC (ω)} are allocated. However, the present invention is not limited to the numerical values described above. This is as described in the embodiment.

  In addition, since discontinuous changes in gain cause deterioration in sound quality, it is preferable to make the gain change smoothly with time, as in the first embodiment, and thus description thereof is omitted. .

  Next, the attenuator (106S1) attenuates and adjusts the microphone sound reception signal (low-pass) by the gain gs1 set (distributed) by the attenuation amount control unit (103), and the attenuator (106Sk) Attenuation adjustment of the microphone sound reception signal (passing through the (k−1) th band) is performed by the gain gsk set (distributed) by the amount control unit (103), and the attenuator (106SN) is controlled by the attenuation amount control unit (103). The microphone sound reception signal (high-pass) is attenuated and adjusted by the gain gsN set (distributed) by.

  The attenuator (106R1) attenuates and adjusts the received signal (low-pass) by the gain gr1 set (distributed) by the attenuation amount control unit (103), and the attenuator (106Rk) is the attenuation amount control unit. The gain (grk) set (distributed) by (103) is used to adjust the attenuation of the received signal (passing through the (k−1) -th band), and the attenuator (106RN) is set (distributed) by the attenuation control unit (103). The received signal (high-pass) is attenuated and adjusted by the gain grN.

  When the set (distributed) gain is 1, the signal is allowed to pass through without being attenuated. That is, when gsj = 1, the microphone sound reception signal (low-pass), the microphone sound reception signal (i-th region passage), and the microphone sound reception signal (high-pass) are output without being attenuated. When grj = 1, the received signal (low-pass), the received signal (i-th pass), and the received signal (high-pass) are output without being attenuated. Thereby, in the case of transmission, the transmission signal can be passed as it is, in the case of reception, the reception signal can be passed as it is, and in the case of single talk, a smooth conversation is possible.

  Next, the adder (107S) receives the microphone sound reception signal (low pass) that has passed through the attenuator (106S1), and the microphone sound reception signal (passed through the (k-1) th pass) that has passed through the attenuator (106Sk). Then, the microphone sound reception signal (high-pass passage) that has passed through the attenuator (106SN) is added to obtain a transmission signal. The transmission signal is transmitted to the other party (speaker D) via a communication device such as a modem.

  The adder (107R) includes a received signal (low-pass) that has passed through the attenuator (106R1), a received signal (passed through the (k−1) -th band) that has passed through the attenuator (106Rk), and an attenuator (106RN). ) Are added and the speaker output signal is output. The speaker output signal is input to the speaker (through D / A conversion or the like) and output as sound from the speaker.

  The above will be described below from another viewpoint, that is, from the viewpoint of the process in which the microphone reception signal and the reception signal are converted into the transmission signal and the speaker output signal, respectively.

  The voice of the speaker C (including noise in the surrounding environment, etc.) is received by the microphone (via A / D conversion) and input to the howling prevention apparatus as a microphone received signal (step S210). . On the other hand, the transmission signal transmitted from the speaker D side is received by a communication device such as a modem and input as a reception signal to the howling prevention device on the speaker C side (step S220).

  The acoustic coupling amount estimation unit (101) obtains the acoustic coupling amount AC (ω) from the microphone sound reception signal and the speaker output signal output from the adder (107R) (step S200). The transmission / reception determination unit (102) determines transmission / reception from the microphone sound reception signal and the reception signal (step S201).

The attenuation amount control unit (103) is divided into three or more (N) frequency bands at a predetermined cutoff frequency from the acoustic coupling amount AC (ω) obtained by the acoustic coupling amount estimation unit (101). Attenuation amount Cj / MAX _ωj {AC (ω)} [j = 1, 2,..., N] in (low frequency band, i-th frequency band, high frequency band) is obtained. Furthermore, the attenuation amount control unit (103) sets the gain gs1 and the attenuator for setting the attenuation amount C1 / MAX _ω1 {AC (ω)} to the attenuator (106S1) from the transmission / reception determination result by the transmission / reception determination unit (102). The gain gr1 set to (106R1) is distributed, the gain gs1 is set to the attenuator (106S1), and the gain gr1 is set to the attenuator (106R1). Further, the attenuation amount control unit (103) determines the attenuation amount Ck / MAX _ωk {AC (ω)} [k = 2, 3,..., N− from the transmission / reception determination result by the transmission / reception determination unit (102). 1] is allocated to the gain gsk set to the attenuator (106Sk) and the gain grk set to the attenuator (106Rk), the gain gsk is set to the attenuator (106Sk), and the gain grk is set to the attenuator (106Rk). Set. Further, the attenuation control unit (103) sets the gain gsN for setting the attenuation CN / MAX _ωN {AC (ω)} in the attenuator (106SN) and the attenuation based on the transmission / reception determination result by the transmission / reception determination unit (102). The gain gsN is set in the attenuator (106SN), and the gain grN is set in the attenuator (106RN) (step S202).

  The low-pass filter (104S) filters the microphone sound reception signal in a predetermined frequency band (step S211a), and the attenuator (106S1) set with the gain gs1 attenuates the filtered microphone sound reception signal. (Step S212a). The band-pass filter (201Sk) filters the microphone sound reception signal in a predetermined frequency band (step S211c), and the attenuator (106Sk) set with the gain gsk then receives the filtered microphone sound reception signal. [K = 2, 3,..., N−1] (step S212c). Further, the high-pass filter (105S) filters the microphone sound reception signal in a predetermined frequency band (step S211b), and then the attenuator (106SN) set with the gain gsN attenuates the filtered microphone sound reception signal. Adjust (step S212b).

  The signals output from the attenuators (106S1) to (106SN) are added by an adder (107S) and output as a transmission signal (step S213). The transmission signal is transmitted to the other party (speaker D) via a communication device such as a modem.

  The low-pass filter (104R) filters the received signal in a predetermined frequency band (step S221a), and then the attenuator (106R1) set with the gain gr1 attenuates the filtered received signal (step S222a). . The band pass filter (201Rk) filters the received signal in a predetermined frequency band (step S221c), and the attenuator (106Rk) set with the gain grk attenuates and adjusts each filtered received signal. [K = 2, 3,..., N−1] (step S222c). Further, the high-pass filter (105R) filters the received signal in a predetermined frequency band (step S221b), and then the attenuator (106RN) set with the gain grN attenuates the filtered received signal (step). S222b).

  The signals output from the attenuators (106R1) to (106RN) are added by the adder (107R) and output as a speaker output signal (step S223). The speaker output signal is input to the speaker (through D / A conversion or the like) and output as sound from the speaker.

  Note that the processes in steps S200 to S202 are repeatedly executed at least as long as a reception signal or a microphone reception signal is input. Of course, the processing of steps S210 to S213 and steps S220 to S223 is repeated for each of the reception signal and the microphone reception signal input to the howling prevention apparatus.

  In the second embodiment, appropriate attenuation can be given to each frequency band of transmission and reception. Further, since the attenuation amount is obtained for each frequency band, it does not affect each other. Therefore, even if the acoustic coupling amount cannot be estimated accurately because no audio signal is included in a certain frequency band, this effect occurs only in the attenuation amount of that frequency band, and the attenuation amount in other frequency bands is appropriate. It can be set to any value. As a result, it is possible to make a call without degrading the two-way call performance.

<Third Embodiment>
Next, a third embodiment of the howling prevention apparatus and method of the present invention will be described.
FIG. 10 is a functional block diagram illustrating how to prevent howling in the howling preventing apparatus according to the third embodiment.
FIG. 11 is a flowchart showing a part of howling prevention processing (setting of attenuation) in the howling prevention apparatus according to the third embodiment.
FIG. 12 is a flowchart showing a part of howling prevention processing (output of a transmission signal) in the howling prevention apparatus according to the third embodiment.
FIG. 13 is a flowchart showing a part of howling prevention processing (output of a speaker output signal) in the howling prevention device according to the third embodiment.

  Descriptions of configurations and functions common to the first embodiment or the second embodiment are omitted, and only parts different from the first embodiment or the second embodiment are described (related to the third embodiment). For an example of the hardware configuration of the howling prevention apparatus, refer to FIG. 1 showing an example of the hardware configuration of the howling prevention apparatus according to the first embodiment.

  The external storage device of the howling prevention apparatus according to the third embodiment includes a program for realizing a filter unit (302S) for filtering a microphone sound reception signal obtained by a microphone, and a filter for the filter unit (302S). A program for realizing a filter coefficient setting unit (301S) for setting a coefficient, a program for realizing a filter unit (302R) for filtering an incoming signal received by a modem, and a filter coefficient for the filter unit (302R) Filter coefficient setting program for realizing the filter coefficient setting unit (301R), an acoustic coupling amount estimation program for obtaining the acoustic coupling amount between the microphone sound reception signal and the output signal input to the speaker, and attenuation from the acoustic coupling amount Attenuation control program for determining the amount to be allocated to the attenuator Grams, handset determination program for the handset determines from the received signal and the microphone received sound signals are stored stored.

  In the howling prevention apparatus according to the third embodiment, each program stored in the external storage device (17) [or ROM, etc.] and data necessary for processing each program are read into the RAM (15) as necessary. Then, the interpretation is executed and processed by the DSP (14). As a result, the DSP (14) realizes predetermined functions (acoustic coupling amount estimation unit, transmission / reception determination unit, attenuation amount control unit, filter coefficient setting unit, filter unit), thereby preventing howling.

Then, next, with reference to FIGS. 10-13, the flow of the howling prevention process in the howling prevention apparatus concerning 3rd Embodiment is demonstrated sequentially.
The howling prevention apparatus according to the third embodiment includes an acoustic coupling amount estimation unit (101), a transmission / reception determination unit (102), an attenuation control unit (103), a filter coefficient setting unit (301R, 301S), and a filter unit ( 302R, 302S).

  In the third embodiment, the low-pass filter, the high-pass filter, the band-pass filter, the attenuator, and the adder in the first embodiment or the second embodiment are each replaced with one filter unit (302R, 302S). It has a configuration. With such a functional configuration, the same processing as in the first embodiment or the second embodiment can be performed with a small amount of calculation.

In the first embodiment or the second embodiment, the input / output relational expressions of the system constituted by the low-pass filter, high-pass filter, band-pass filter, attenuator and adder are the expressions (14) and (15). It is represented by

Here, In _S (ω) represents a microphone sound reception signal, In _R (ω) represents a reception signal, Out _S (ω) represents a transmission signal, and Out _R (ω) represents a speaker output signal. Further, gsj and grj represent gains distributed by the attenuation amount control unit (102), and fsj (ω) and frj (ω) are filters for filtering in the j-th frequency band (low-pass filter, band-pass filter, Represents the transfer function of the high-pass filter. These transfer functions are assumed to be known and incorporated in the filter coefficient setting program described above.

If the microphone sound reception signal In _S (ω) is filtered by a filter unit using flt _S (ω) calculated by Expression (16) as a filter coefficient, the same output Out as that expressed by Expression (14) is obtained. _S (ω) is obtained (see equation (17)), and the received signal In _R (ω) is filtered by a filter unit using flt _R (ω) calculated by equation (18) as a filter coefficient. The same output Out _R (ω) as expressed by (15) is obtained (see equation (19)).

Therefore, the same filter as the first embodiment or the second embodiment is used by the filter unit having the filter coefficients of flt _S (ω) and flt _R (ω) calculated by the equations (16) and (18). Outputs Out _S (ω) and Out _R (ω) can be obtained.

  The following description will be made from the viewpoint of the process in which the microphone sound reception signal and the reception signal are converted into the transmission signal and the speaker output signal, respectively.

  The voice of the speaker C (including noise in the surrounding environment, etc.) is received by the microphone (via A / D conversion) and input to the howling prevention device as a microphone received signal (step S310). . On the other hand, the transmission signal transmitted from the speaker D side is received by a communication device such as a modem and inputted as a reception signal to the howling prevention device on the speaker C side (step S320).

  The acoustic coupling amount estimation unit (101) obtains the acoustic coupling amount AC (ω) from the microphone sound reception signal and the speaker output signal output from the filter unit (302R) (step S300). The transmission / reception determination unit (102) determines transmission / reception from the microphone sound reception signal and the reception signal (step S301).

The attenuation amount control unit (103) has a plurality of (two in the case corresponding to the first embodiment) at a predetermined cutoff frequency from the acoustic coupling amount AC (ω) obtained by the acoustic coupling amount estimation unit (101). In the case of corresponding to the second embodiment, each frequency band divided into N [≧ 3] (low frequency band and high frequency band in the case of corresponding to the first embodiment, the second embodiment) In the corresponding case, the attenuation Cj / MAX _ωj {AC (ω)} [j = 1, 2,..., Y in the low frequency band, the i th frequency band, and the high frequency band), where Y is an integer of 2 or more And ]. Further, the attenuation control unit (103) determines the attenuation Cj / MAX _ωj {AC (ω)} [j = 1, 2,... For each frequency band from the transmission / reception determination result by the transmission / reception determination unit (102). , Y] are distributed to the gain gsj and the gain grj (step S302).

  The filter coefficient setting unit (301S) includes a gain gsj [j = 1, 2,..., Y] distributed by the attenuation amount control unit (103) and a known filter (described in the first and second embodiments). A filter coefficient (see equation (16)) is obtained from the transfer function of the low-pass filter, high-pass filter, and band-pass filter that performs filtering in each frequency band, and this filter coefficient is set as the filter coefficient in the filter unit (302S) ( Step S303).

  The filter coefficient setting unit (301R) includes a gain grj [j = 1, 2,..., Y] distributed by the attenuation amount control unit (103) and a known filter (described in the first and second embodiments). A filter coefficient (see equation (18)) is obtained from the transfer function of the low-pass filter, high-pass filter, and band-pass filter that performs filtering in each frequency band, and this filter coefficient is set as the filter coefficient in the filter unit (302R) ( Step S303).

  The filter unit (302S) filters the microphone sound reception signal and outputs a transmission signal (step S311). The transmission signal is transmitted to the other party (speaker D) via a communication device such as a modem.

  The filter unit (302R) filters the reception signal and outputs a speaker output signal (step S321). The speaker output signal is input to the speaker (through D / A conversion or the like) and output as sound from the speaker.

  Note that the processing in steps S300 to S303 is repeatedly executed at least as long as a reception signal or a microphone reception signal is input. Of course, the processing of step S310, step S311, step S320, and step S321 is repeated for each of the reception signal and microphone reception signal input to the howling prevention apparatus.

  Since the filter coefficient calculation according to Expression (16) and Expression (18) does not require a convolution operation, the filter coefficient can be calculated with a small amount of calculation. From this, according to 3rd Embodiment, the same effect as 1st Embodiment or 2nd Embodiment can be acquired with little calculation amount.

<Fourth Embodiment>
In the fourth embodiment of the present invention, in the attenuation control unit (103) of the first to third embodiments, the attenuation Cj / MAX _ωj {AC (ω)} [j = 1, 2 _,. .., Y] is determined so that gs1 · gr1 ≧ gs2 · gr2 ≧... ≧ gsY · grY, and the attenuation amount becomes smaller (the gain becomes larger) in the lower frequency band (attenuation amount). Cj / MAX _ωj {AC (ω)} is not the attenuation amount of the attenuator The attenuation amount Z of the attenuator is determined by the gain gsj and the gain grj to which the attenuation amount Cj / MAX _ωj {AC (ω)} is allocated. Note that it is determined.) The distribution of the gain gsj on the transmission side and the gain grj on the reception side is the same as that in the first to third embodiments, and thus the description thereof is omitted.

  When the attenuation amount in the lower frequency band is large and the attenuation amount in the higher frequency band is small, only the audio signal in the higher frequency band is communicated, which may cause deterioration in sound quality. Therefore, in the fourth embodiment, in order to prevent this, the attenuation amount in the higher frequency band is set larger than the attenuation amount in the lower frequency band.

  As described above, according to the fourth embodiment, in addition to the effects of the first to third embodiments, it is possible to prevent a sound signal of a lower frequency band from being communicated and causing deterioration in sound quality.

<Fifth Embodiment>
Next, a fifth embodiment of the howling prevention apparatus and method of the present invention will be described.
FIG. 14 is a functional block diagram illustrating how to prevent howling in the howling preventing apparatus according to the fifth embodiment.
FIG. 15 is a flowchart showing a part of howling prevention processing (output of a transmission signal) in the howling prevention apparatus according to the fifth embodiment.
FIG. 16 is a flowchart illustrating a part of howling prevention processing (output of a speaker output signal) in the howling prevention device according to the fifth embodiment.

  The description of the configurations and functions common to the first to fourth embodiments will be omitted, and only the parts different from the first to fourth embodiments will be described (the hardware of the howling prevention apparatus according to the fifth embodiment). 1 is a hardware configuration example of a howling prevention apparatus according to the first embodiment, and a part of howling prevention processing (attenuation amount) in the howling prevention apparatus according to the fifth embodiment. (Refer to FIG. 7 showing a flowchart showing a part of howling prevention processing (setting of attenuation) in the howling prevention apparatus according to the second embodiment.)

  The external storage device of the howling prevention apparatus according to the fifth embodiment includes a program for performing a short-time Fourier transform of a microphone sound reception signal obtained by a microphone, and a predetermined frequency of the microphone sound reception signal subjected to the short-time Fourier transform. Attenuator (106S1 to 106SN) for adjusting the attenuation of the band [in the fifth embodiment, N is an integer of 2 or more. ], A program for performing a short-time inverse Fourier transform on each microphone received signal subjected to attenuation adjustment, a program for performing a short-time Fourier transform on the received signal received by the receiving device, and a short-time Fourier transform Program for realizing an attenuator (106R1 to 106RN) for attenuation adjustment of a predetermined frequency band of the received reception signal, a program for short-time inverse Fourier transform of each reception signal subjected to attenuation adjustment, and microphone reception signal Coupling amount estimation program for determining the amount of acoustic coupling between the sound and the output signal input to the speaker, attenuation control program for determining the amount of attenuation from the amount of acoustic coupling and allocating it to the attenuator, received signal and microphone reception A transmission / reception determination program for determining transmission / reception from the signal is stored and stored.

  In the howling prevention apparatus according to the fifth embodiment, each program stored in the external storage device (17) [or ROM, etc.] and data necessary for processing each program are read into the RAM (15) as necessary. Then, the interpretation is executed and processed by the DSP (14). As a result, the DSP (14) realizes predetermined functions (acoustic coupling amount estimation unit, transmission / reception determination unit, attenuation control unit, short-time Fourier transform unit, attenuator, short-time inverse Fourier transform unit), Howling can be prevented.

Then, next, with reference to FIGS. 14-16, the flow of the howling prevention process in the howling prevention apparatus concerning 5th Embodiment is demonstrated sequentially.
The howling prevention apparatus of the present embodiment includes an acoustic coupling amount estimation unit (101), a transmission / reception determination unit (102), an attenuation control unit (103), a reception-side attenuator (106R1 to 106RN), and a transmission-side attenuation. Functions (106S1-106SN), short-time Fourier transform (401R, 401S) and short-time inverse Fourier transform unit (402R, 402S).

  The short-time Fourier transform unit (401S) performs a short-time Fourier transform on the microphone sound reception signal and outputs a frequency-domain microphone sound reception signal. Each attenuator (106S1 to 106SN) is a frequency band obtained by dividing the frequency band in the short-time Fourier transform into N frequency bands from the frequency domain microphone sound reception signal output from the short-time Fourier transform unit (401S). The frequency domain microphone sound reception signal is attenuated. That is, for example, the j-th attenuator (106Sj) is the j-th frequency band in the short-time Fourier transform divided into N out of the frequency domain microphone received signals output from the short-time Fourier transform unit (401S). Attenuation adjustment is performed on the frequency domain microphone sound reception signal in the frequency band. The attenuation amount control unit (103) obtains and distributes the gains gsj and grj of the respective attenuators (106S1 to 106SN) in the same manner as described in the second embodiment, and thus description thereof is omitted.

  The short-time inverse Fourier transform unit (402S) obtains a transmission signal by performing a short-time inverse Fourier transform on each frequency domain microphone received signal that has passed through each attenuator (106S1 to 106SN).

  What has been described above also applies to the short-time Fourier transform unit (401R), the attenuators (106R1 to 106RN), and the short-time Fourier transform unit (402R) on the receiver side.

  This will be described below from the viewpoint of the process in which the microphone reception signal and the reception signal are converted into the transmission signal and the speaker output signal, respectively.

  The voice of speaker C (including noise in the surrounding environment, etc.) is received by the microphone (via A / D conversion) and input to the howling prevention apparatus as a microphone received signal (step S400a). . On the other hand, the transmission signal transmitted from the speaker D side is received by a communication device such as a modem and is input as a reception signal to the howling prevention device on the speaker C side (step S400b).

  The acoustic coupling amount estimation unit (101) obtains the acoustic coupling amount AC (ω) from the microphone sound reception signal and the speaker output signal output from the short-time inverse Fourier transform unit (402R) (see step S200). The transmission / reception determination unit (102) determines transmission / reception from the microphone sound reception signal and the reception signal (see step S201).

The attenuation control unit (103) is configured to reduce the attenuation Cj / MAX for each frequency band in the short-time Fourier transform divided into N from the acoustic coupling amount AC (ω) obtained by the acoustic coupling amount estimation unit (101). _ωj {AC (ω)} [j = 1, 2,..., N, where N is an integer of 2 or more. ]. Further, the attenuation control unit (103) sets the attenuation Cj / MAX _ωj {AC (ω)} to the attenuator (106Sj) for each frequency band from the transmission / reception determination result by the transmission / reception determination unit (102). The gain gsj and the gain grj set in the attenuator (106Rj) are distributed, the gain gsj is set in the attenuator (106Sj), and the gain grj is set in the attenuator (106Rj) (see step S202).

  The short-time Fourier transform unit (401S) performs short-time Fourier transform on the microphone sound reception signal and outputs a frequency domain microphone sound reception signal (step S410a). Next, each attenuator (106Sj) attenuates and adjusts the frequency domain microphone sound reception signal in the jth frequency band (step S402a).

  The short-time inverse Fourier transform unit (402S) obtains a transmission signal by performing a short-time inverse Fourier transform on each frequency domain microphone received signal that has passed through each attenuator (106S1 to 106SN) (step S403a). The transmission signal is transmitted to the other party (speaker D) via a communication device such as a modem.

  The short-time Fourier transform unit (401R) performs short-time Fourier transform on the received signal and outputs a frequency domain received signal (step S401b). Next, each attenuator (106Rj) performs attenuation adjustment of the frequency domain received signal in the jth frequency band (step S402b).

  The short-time inverse Fourier transform unit (402R) performs short-time inverse Fourier transform on each frequency domain received signal that has passed through each attenuator (106R1 to 106RN) to obtain a speaker output signal (step S403b). The speaker output signal is input to the speaker (through D / A conversion or the like) and output as sound from the speaker.

  Note that the processing in steps S200 to S202 in the fifth embodiment is repeatedly executed at least as long as a reception signal or a microphone reception signal is input. Of course, the processing of step S400a to step S403a and step S400b to step S403b is repeated for each of the reception signal and microphone reception signal input to the howling prevention apparatus.

  When the number of band divisions is large, it is possible to reduce the amount of calculation by band division by short-time Fourier transform than by using a time domain filter. As described above, according to the fifth embodiment, in addition to the effects of the first to fourth embodiments, it is possible to implement with a low calculation amount even when the number of band divisions is large.

<Sixth Embodiment>
17 and 18 are functional block diagrams of a howling prevention apparatus according to the sixth embodiment of the present invention.
The howling prevention apparatus according to the sixth embodiment corresponds to the howling prevention unit 501 of the above first to fifth embodiments [the functional component part surrounded by the broken line in FIGS. 2, 6, 10 and 14. ] Is a functional configuration (see FIG. 17) provided with an echo canceller unit 502 in the previous stage (see FIG. 17), or a functional configuration (see FIG. 18) provided with an echo canceller unit 502 and a microphone array unit 503 in the previous stage of the howling prevention unit 501.

  The echo canceller unit 502 suppresses echo signals using an adaptive filter or the like, and the microphone array unit 503 performs noise suppression, volume adjustment, echo suppression, and the like. Thereby, the amount of acoustic coupling in the howling prevention unit is reduced, and the two-way call performance is further improved.

  In addition, the howling prevention apparatus / method according to the present invention is not limited to the above-described embodiment, and can be appropriately changed without departing from the gist of the present invention. In addition, the processing described in the above howling prevention apparatus / method is not only executed in time series in the order described, but also executed in parallel or individually as required by the processing capability of the apparatus that executes the processing. It is good.

  When the processing function in the howling prevention apparatus is realized by a computer, the processing content of the function that the howling prevention apparatus should have is described by a program. By executing this program on a computer, the processing function of the howling prevention device is realized on the computer.

  The program describing the processing contents can be recorded on a computer-readable recording medium. As the computer-readable recording medium, for example, any recording medium such as a magnetic recording device, an optical disk, a magneto-optical recording medium, and a semiconductor memory may be used. Specifically, for example, as a magnetic recording device, a hard disk device, a flexible disk, a magnetic tape or the like, and as an optical disk, a DVD (Digital Versatile Disc), a DVD-RAM (Random Access Memory), a CD-ROM (Compact Disc Read Only). Memory), CD-R (Recordable) / RW (ReWritable), etc., magneto-optical recording medium, MO (Magneto-Optical disc), etc., semiconductor memory, EEP-ROM (Electronically Erasable and Programmable-Read Only Memory), etc. Can be used.

  The program is distributed by selling, transferring, or lending a portable recording medium such as a DVD or CD-ROM in which the program is recorded. Furthermore, the program may be distributed by storing the program in a storage device of the server computer and transferring the program from the server computer to another computer via a network.

  A computer that executes such a program first stores, for example, a program recorded on a portable recording medium or a program transferred from a server computer in its own storage device. When executing the process, the computer reads a program stored in its own recording medium and executes a process according to the read program. As another execution form of the program, the computer may directly read the program from a portable recording medium and execute processing according to the program, and the program is transferred from the server computer to the computer. Each time, the processing according to the received program may be executed sequentially. Also, the program is not transferred from the server computer to the computer, and the above-described processing is executed by a so-called ASP (Application Service Provider) type service that realizes the processing function only by the execution instruction and result acquisition. It is good. Note that the program in this embodiment includes information that is used for processing by an electronic computer and that conforms to the program (data that is not a direct command to the computer but has a property that defines the processing of the computer).

  In this embodiment, the howling prevention apparatus is configured by executing a predetermined program on a computer. However, at least a part of these processing contents may be realized by hardware.

  The present invention is useful for hands-free calls such as a TV conference and an audio conference.

The block diagram which illustrated the hardware constitutions of the howling prevention apparatus (1) concerning 1st Embodiment. The functional block diagram which illustrates the processing function of howling prevention in the howling prevention apparatus (1) concerning a 1st embodiment. The flowchart which shows a part (setting of attenuation amount) of the howling prevention process in the howling prevention apparatus (1) concerning 1st Embodiment. The flowchart which shows a part of howling prevention process (output of a transmission signal) in the howling prevention apparatus (1) concerning 1st Embodiment. The flowchart which shows a part of howling prevention process (output of a speaker output signal) in the howling prevention apparatus (1) concerning 1st Embodiment. The functional block diagram which illustrates the processing function of howling prevention in the howling prevention device concerning a 2nd embodiment. 10 is a flowchart showing a part of howling prevention processing (setting of attenuation amount) in the howling prevention device according to the second embodiment. 10 is a flowchart showing a part of howling prevention processing (output of a transmission signal) in the howling prevention device according to the second embodiment. 9 is a flowchart showing a part of howling prevention processing (output of a speaker output signal) in the howling prevention device according to the second embodiment. The functional block diagram which illustrates the processing function of howling prevention in the howling prevention device concerning a 3rd embodiment. 10 is a flowchart showing a part of howling prevention processing (setting of attenuation) in a howling prevention apparatus according to the third embodiment. The flowchart which shows a part (output of a transmission signal) of the howling prevention process in the howling prevention apparatus concerning 3rd Embodiment. 12 is a flowchart showing a part of howling prevention processing (output of a speaker output signal) in a howling prevention device according to the third embodiment. The functional block diagram which illustrates the processing function of howling prevention in the howling prevention device concerning a 5th embodiment. The flowchart which shows a part of howling prevention processing (output of a transmission signal) in the howling prevention apparatus concerning 5th Embodiment. The flowchart which shows a part of howling prevention processing (output of a speaker output signal) in the howling prevention apparatus concerning 5th Embodiment. The functional block diagram of the howling prevention apparatus concerning 6th Embodiment. The functional block diagram of the howling prevention apparatus concerning 6th Embodiment. The figure explaining the conventional howling prevention apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Acoustic coupling amount estimation part 102 Transmission / reception determination part 103 Attenuation amount control part 104R Low-pass filter 104S on reception side Low-pass filter 105R on transmission side High-pass filter 105S on reception side High-pass filter 106R1-106RN on transmission side Attenuator on reception side 106S1 to 106SN Transmission side attenuator 107R Reception side adder 107S Transmission side adder 201R1 to 201RM Reception side bandpass filter 201S1 to 201SM Transmission side bandpass filter 301R Reception side filter coefficient setting unit 301S Transmission-side filter coefficient setting unit 302R Reception-side filter unit 302S Transmission-side filter unit 401R Reception-side short-time Fourier transform unit 401S Transmission-side short-time Fourier transform unit 402R Reception-side short-time inverse Fourier Conversion unit 402S Short-time reverse on the transmission side Fourier transform unit 501 Howling prevention unit 502 of the present invention Echo canceller unit 503 Microphone array unit

Claims (12)

An input transmission conversion step of converting an input signal received by the sound receiving device into a transmission signal;
A reception output conversion step of converting a reception signal received by the reception device into an output signal input to the output device;
An acoustic coupling amount estimating step for obtaining an acoustic coupling amount that is an amplitude of a transfer function between the output device and the sound receiving device from the input signal and the output signal;
A transmission / reception determination step for determining whether the transmission / reception determination means of the howling prevention apparatus is a transmission signal or a reception signal from the input signal and the reception signal;
The attenuation amount control means of the howling prevention device uses an attenuation amount that does not generate howling from the acoustic coupling amount obtained in the acoustic coupling amount estimation step as a frequency band below a predetermined frequency (hereinafter referred to as “low frequency band”) and. Attenuation amount obtained for each frequency band equal to or higher than the frequency (hereinafter referred to as “high frequency band”), and each attenuation amount is distributed to each attenuation means of the howling prevention device based on the transmission / reception determination result in the transmission / reception determination step. Control steps,
The input transmission conversion step is
An input signal low-pass step of the input signal low-pass means of the howling prevention device for filtering the input signal and passing the input signal in the low frequency band;
An input signal high-pass step in which the input signal high-pass means of the howling prevention device filters the input signal and passes the input signal in the high frequency band, and
A low-frequency input signal attenuation step in which the low-frequency input signal attenuation means of the howling prevention device attenuates the low-frequency input signal passed in the input signal low-frequency pass step based on the attenuation amount allocated in the attenuation amount control step;
A high-frequency input signal attenuation step in which the high-frequency input signal attenuation means of the howling prevention device attenuates the high-frequency input signal passed in the input signal high-frequency pass step based on the attenuation amount allocated in the attenuation amount control step;
The input signal adding means of the howling prevention device adds the low-frequency input signal whose attenuation is adjusted in the low-frequency input signal attenuation step and the high-frequency input signal whose attenuation is adjusted in the high-frequency input signal attenuation step to obtain the transmission signal. An input signal adding step to obtain,
The received output conversion step includes
The reception signal low-pass means of the reception signal low-pass means of the howling prevention device filters the reception signal and passes the reception signal of the low frequency band, and
The reception signal high-pass means of the reception signal high-pass means of the howling prevention device filters the reception signal and passes the reception signal in the high frequency band, and
A low-frequency received signal attenuation step in which the low-frequency received signal attenuation means of the howling prevention device attenuates and adjusts the low-frequency received signal passed in the received signal low-frequency pass step based on the attenuation amount allocated in the attenuation amount control step;
A high-frequency received signal attenuation step in which the high-frequency received signal attenuation means of the howling prevention device attenuates and adjusts the high-frequency received signal passed in the received signal high-frequency pass step based on the attenuation amount allocated in the attenuation amount control step;
The reception signal adding means of the howling prevention device adds the low-frequency received signal whose attenuation is adjusted in the low-frequency received signal attenuation step and the high-frequency received signal whose attenuation is adjusted in the high-frequency received signal attenuation step to obtain an output signal. A receiving signal adding step for obtaining a howling prevention method. An input transmission conversion step of converting an input signal received by the sound receiving device into a transmission signal;
A reception output conversion step of converting a reception signal received by the reception device into an output signal input to the output device;
An acoustic coupling amount estimating step for obtaining an acoustic coupling amount that is an amplitude of a transfer function between the output device and the sound receiving device from the input signal and the output signal;
A transmission / reception determination step for determining whether the transmission / reception determination means of the howling prevention apparatus is a transmission signal or a reception signal from the input signal and the reception signal;
The attenuation amount control means of the howling prevention apparatus uses an attenuation amount that does not generate howling from the acoustic coupling amount obtained in the acoustic coupling amount estimation step as a frequency band of frequency ω ₁ or less (hereinafter referred to as “low frequency band”). i = 1, 2,..., M (M is a natural number of 1 or more), a frequency band of frequency ω _i or more and ω _{i + 1} or less (hereinafter referred to as “i-th frequency band”), a frequency of frequency ω _{M + 1} or more. An attenuation amount control step for determining each attenuation amount for each band (hereinafter referred to as “high frequency band”) and allocating each attenuation amount to each attenuation means of the howling prevention device based on the transmission / reception determination result in the transmission / reception determination step. And
The input transmission conversion step is
An input signal low-pass step of the input signal low-pass means of the howling prevention device for filtering the input signal and passing the input signal in the low frequency band;
An input signal i-th frequency band passing step in which each of the M input signal i-th frequency band passing means of the howling prevention device filters the input signal and passes the input signal in the i-th frequency band;
The input signal high-pass means of the input signal high-pass means of the howling prevention device filters the input signal and passes the input signal in the high frequency band, and
A low-frequency input signal attenuation step in which the low-frequency input signal attenuation means of the howling prevention device attenuates the low-frequency input signal passed in the input signal low-frequency pass step based on the attenuation amount allocated in the attenuation amount control step;
The i-th frequency band input signal attenuating means of the howling prevention apparatus adjusts the attenuation of the i-th frequency band input signal passed in the input signal i-th frequency band passing step based on the attenuation amount allocated in the attenuation amount control step. A frequency band input signal attenuation step;
A high-frequency input signal attenuation step in which the high-frequency input signal attenuation means of the howling prevention device attenuates the high-frequency input signal passed in the input signal high-frequency pass step based on the attenuation amount allocated in the attenuation amount control step;
The input signal adding means of the howling prevention device includes a low-frequency input signal that is attenuated in the low-frequency input signal attenuation step, an i-th frequency band input signal that is attenuated in the i-th frequency band input signal attenuation step, and a high-frequency input. It consists of an input signal addition step that obtains a transmission signal by adding the high-frequency input signal whose attenuation has been adjusted in the signal attenuation step,
The received output conversion step includes
The reception signal low-pass means of the reception signal low-pass means of the howling prevention device filters the reception signal and passes the reception signal of the low frequency band, and
A received signal i-th frequency band passing step for filtering the received signal and passing the received signal in the i-th frequency band by the M received signal i-th frequency band passing means of the howling prevention apparatus;
The reception signal high-pass means of the reception signal high-pass means of the howling prevention device filters the reception signal and passes the reception signal in the high frequency band, and
A low-frequency received signal attenuation step in which the low-frequency received signal attenuation means of the howling prevention device attenuates and adjusts the low-frequency received signal passed in the received signal low-frequency pass step based on the attenuation amount allocated in the attenuation amount control step;
The i th frequency band received signal attenuating means of the howling prevention device adjusts the attenuation of the i th frequency band received signal passed in the received signal i th frequency band passing step based on the attenuation amount allocated in the attenuation amount control step. A frequency band received signal attenuation step;
A high-frequency received signal attenuation step in which the high-frequency received signal attenuation means of the howling prevention device attenuates and adjusts the high-frequency received signal passed in the received signal high-frequency pass step based on the attenuation amount allocated in the attenuation amount control step;
The reception signal adding means of the howling prevention apparatus includes a low-frequency reception signal that is attenuated in the low-frequency reception signal attenuation step, an i-th frequency band reception signal that is attenuated in the i-th frequency band reception signal attenuation step, and a high-frequency reception. A method for preventing howling, comprising: a received signal adding step of adding an output signal obtained by adding a high frequency received signal whose attenuation is adjusted in the signal attenuating step. Input transmission conversion step is
An input signal filter coefficient setting step of the input signal filter coefficient setting means of the howling prevention apparatus for obtaining a filter coefficient in which the attenuation amount for each frequency band allocated in the attenuation amount control step becomes a frequency characteristic;
The input signal filtering means of the howling prevention device comprises an input signal filtering step of filtering the input signal by the filter coefficient obtained in the input signal filter coefficient setting step,
Receive output conversion step
A reception signal filter coefficient setting step of the reception signal filter coefficient setting means of the howling prevention device for obtaining a filter coefficient in which the attenuation amount for each frequency band allocated by the attenuation amount control means becomes a frequency characteristic;
The reception signal filtering means of the howling prevention apparatus comprises a reception signal filtering step of filtering the reception signal by the filter coefficient obtained in the reception signal filter coefficient setting step. Howling prevention method. The attenuation control step is
The attenuation control means of the howling prevention apparatus sets each frequency band to B1, B2,..., BN (where N is an integer of 2 or more) from the lower frequency band, and each frequency band B1, B2,. , BN, the attenuation amounts corresponding to G1, G2,..., GN are obtained, and the attenuation amounts G1, G2,. 4. The howling prevention method according to claim 1, wherein distribution is made to each attenuation means of the howling prevention device based on the transmission / reception judgment result obtained in the transmission / reception judgment step. Input transmission conversion step is
The input signal short-time Fourier transform means of the input signal short-time Fourier transform step of the input signal short-time Fourier transform of the howling prevention device,
Each STFT input signal attenuation means of the howling prevention device attenuates and adjusts the STFT input signal output in the input signal short-time Fourier transform step based on the attenuation amount allocated in the attenuation amount control step for each frequency band. STFT input signal attenuation step;
The input signal short-time inverse Fourier transform means of the howling prevention device comprises an input signal short-time inverse Fourier transform step for performing a short-time inverse Fourier transform on the ATT input signal output in the STFT input signal attenuation step,
Receive output conversion step
The reception signal short-time Fourier transform means of the howling prevention device receives the short-time Fourier transform step for short-time Fourier transform of the reception signal, and
Each STFT reception signal attenuating means of the howling prevention device attenuates the STFT reception signal output in the reception signal short-time Fourier transform step based on the attenuation amount allocated in the attenuation amount control step for each frequency band. STFT received signal attenuation step;
The reception signal short-time inverse Fourier transforming means of the howling prevention device comprises a reception signal short-time inverse Fourier transform step for performing short-time inverse Fourier transform on the ATT reception signal output in the STFT reception signal attenuation step, The howling prevention method according to claim 1, claim 2, or claim 4. Input transmission conversion means for converting an input signal received by the sound receiving device into a transmission signal;
Reception output conversion means for converting a reception signal received by the reception device into an output signal input to the output device;
An acoustic coupling amount estimating means for obtaining an acoustic coupling amount which is an amplitude of a transfer function between the output device and the sound receiving device from the input signal and the output signal;
A transmission / reception determination means for determining whether a transmission signal or a reception signal from an input signal and a reception signal;
The attenuation that does not cause howling from the acoustic coupling amount obtained by the acoustic coupling amount estimation means is defined as a frequency band below a predetermined frequency (hereinafter referred to as “low frequency band”) and a frequency band above the frequency (hereinafter referred to as “ Attenuation amount control means for allocating the respective attenuation amounts to the input transmission conversion means and the reception output conversion means based on the transmission / reception determination result obtained by the transmission / reception determination means. ,
Input transmission conversion means
An input signal low-pass means for filtering the input signal and passing the input signal in the low frequency band;
Input signal high-pass means for filtering the input signal and passing the input signal in the high frequency band;
Low-frequency input signal attenuation means for adjusting the attenuation of the low-frequency input signal passed by the low-frequency input means based on the attenuation amount allocated by the attenuation amount control means;
High-frequency input signal attenuation means for adjusting the attenuation of the high-frequency input signal passed by the input signal high-frequency pass means based on the attenuation amount allocated by the attenuation amount control means;
The low-frequency input signal attenuated by the low-frequency input signal attenuating means and the high-frequency input signal attenuated by the high-frequency input signal attenuating means are added to obtain the transmission signal, and the input signal adding means is configured to obtain a transmission signal.
Reception output conversion means
Received signal low-pass means for filtering the received signal and passing the received signal in the low frequency band,
Received signal high-pass means for filtering the received signal and passing the received signal in the high frequency band;
Low-frequency received signal attenuation means for adjusting the attenuation of the low-frequency received signal passed by the received signal low-frequency pass means based on the attenuation amount allocated by the attenuation amount control means;
High-frequency received signal attenuation means for adjusting the attenuation of the high-frequency received signal passed by the received signal high-frequency passing means based on the attenuation amount distributed by the attenuation amount control means;
A reception signal adding means for adding an output signal obtained by adding the low frequency reception signal attenuated by the low frequency reception signal attenuation means and the high frequency reception signal adjusted by the high frequency reception signal attenuation means. A howling prevention device characterized by that. Input transmission conversion means for converting an input signal received by the sound receiving device into a transmission signal;
Reception output conversion means for converting a reception signal received by the reception device into an output signal input to the output device;
An acoustic coupling amount estimating means for obtaining an acoustic coupling amount which is an amplitude of a transfer function between the output device and the sound receiving device from the input signal and the output signal;
A transmission / reception determination means for determining whether a transmission signal or a reception signal from an input signal and a reception signal;
Attenuation amount that does not produce howling from the acoustic coupling amount obtained by the acoustic coupling amount estimating means is a frequency band of frequency ω ₁ or less (hereinafter referred to as “low frequency band”), i = 1, 2,. M (M is a natural number of 1 or more), a frequency band of frequency ω _i or more and ω _{i + 1} or less (hereinafter referred to as “i-th frequency band”), a frequency band of frequency ω _{M + 1} or more (hereinafter referred to as “high frequency band”) .) And attenuation amount control means for allocating each attenuation amount to the input transmission conversion means and the reception output conversion means based on the transmission / reception determination result obtained by the transmission / reception determination means,
Input transmission conversion means
An input signal low-pass means for filtering the input signal and passing the input signal in the low frequency band;
M input signal i th frequency band passing means for filtering the input signal and passing the input signal of the i th frequency band;
Input signal high-pass means for filtering the input signal and passing the input signal in the high frequency band;
Low-frequency input signal attenuation means for adjusting the attenuation of the low-frequency input signal passed by the input signal low-frequency means based on the attenuation amount allocated by the attenuation amount control means;
M number of i th frequency band input signal attenuating means for adjusting the attenuation of the i th frequency band input signal passed by each i th frequency band passing means based on the amount of attenuation allocated by the amount of attenuation control means;
High-frequency input signal attenuation means for adjusting the attenuation of the high-frequency input signal passed by the input signal high-frequency pass means based on the attenuation amount allocated by the attenuation amount control means;
Low-frequency input signal attenuated by low-frequency input signal attenuating means, i-th frequency band input signal attenuated by each i-th frequency band input signal attenuating means, and high-frequency attenuated by high frequency input signal attenuating means The input signal adding means for adding the input signal to obtain the transmission signal,
Reception output conversion means
Received signal low-pass means for filtering the received signal and passing the received signal in the low frequency band,
M received signal i th frequency band passing means for filtering the received signal to pass the received signal in the i th frequency band;
Received signal high-pass means for filtering the received signal and passing the received signal in the high frequency band;
Low-frequency received signal attenuation means for adjusting the attenuation of the low-frequency received signal passed by the received signal low-frequency pass means based on the attenuation amount allocated by the attenuation amount control means;
M number of i th frequency band received signal attenuating means for adjusting the attenuation of the i th frequency band received signal passed by each i th frequency band passing means based on the amount of attenuation allocated by the amount of attenuation control means;
High-frequency received signal attenuation means for adjusting the attenuation of the high-frequency received signal passed by the received signal high-frequency passing means based on the attenuation amount distributed by the attenuation amount control means;
Low-frequency received signal attenuated by low-frequency received signal attenuating means, i-th frequency band received signal adjusted for attenuation by each i-th frequency band received signal attenuating means, and high-frequency adjusted by high-frequency received signal attenuating means A howling prevention apparatus comprising reception signal adding means for adding an output signal to obtain an output signal. Input transmission conversion means
Input signal filter coefficient setting means for obtaining a filter coefficient in which the attenuation amount for each frequency band allocated by the attenuation amount control means is a frequency characteristic;
The input signal filtering means for filtering the input signal by the filter coefficient obtained by the input signal filter coefficient setting means,
Reception output conversion means
A received signal filter coefficient setting means for obtaining a filter coefficient in which the attenuation amount for each frequency band allocated by the attenuation amount control means is a frequency characteristic;
8. The howling prevention apparatus according to claim 6, further comprising: reception signal filtering means for filtering the reception signal by the filter coefficient obtained by the reception signal filter coefficient setting means. The attenuation control means is
Each frequency band is set to B1, B2,..., BN (where N is an integer equal to or greater than 2) from a low frequency band, and the attenuation corresponding to each frequency band B1, B2,. If G2,..., GN, the attenuation amount satisfying G1 ≧ G2 ≧... ≧ GN is obtained, and the attenuation amounts G1, G2,. 9. The howling prevention apparatus according to claim 6, wherein the howling prevention device is distributed to the input transmission conversion means and the reception output conversion means on the basis of. Input transmission conversion means
Input signal short-time Fourier transform means for short-time Fourier transform of the input signal;
A plurality of STFT input signal attenuating means for adjusting the STFT input signal output by the input signal short-time Fourier transform means for each frequency band based on the attenuation distributed by the attenuation control means;
An input signal short time inverse Fourier transform means for performing a short time inverse Fourier transform on the ATT input signal output by each STFT input signal attenuation means,
Reception output conversion means
Received signal short-time Fourier transform means for short-time Fourier transform of the received signal;
A plurality of STFT received signal attenuating means for adjusting the STFT received signal output by the received signal short-time Fourier transform means based on the amount of attenuation allocated by the attenuation amount controlling means for each frequency band;
The reception signal short-time inverse Fourier transform means for performing a short-time inverse Fourier transform on the ATT reception signal output by each STFT reception signal attenuating means comprises any one of claims 6, 7, and 9. The howling prevention apparatus of crab.       A howling prevention program for causing a computer to execute the howling prevention method according to any one of claims 1 to 5.       The computer-readable recording medium which recorded the howling prevention program of Claim 11.

Images