JP2007318274A - Sound emission/pickup apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To suppress sneak sound more effectively. <P>SOLUTION: A first filter 203 delivers the frequency characteristics of an inputted sound pickup beam signal to an echo cancel circuit 200 while correcting such that the sneak sound has a uniform signal level over the entire frequency band. The echo cancel circuit 200 generates a para-recurrence sound signal from an input sound signal of sound pickup signal and subtracts it from a sound pickup beam signal corrected through the first filter 203. Since the frequency characteristics of an sneak signal are substantially uniform over the entire band in the first filter 203, the sneak sound signal and the para-recurrence sound signal included in the corrected sound pickup beam signal have a substantially identical frequency spectrum. Consequently, the sneak sound is removed effectively by the echo cancel circuit 200. A second filter 204 readjusts the frequency characteristics of utterance sound from a person of utterance included in the sound pickup beam signal and varied along with the sneak sound to the original state before delivering an output sound signal. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a sound emission and collection device used for audio conferences or the like performed between a plurality of points via a network or the like, and more particularly to a sound emission and collection device having an echo cancellation function.

  2. Description of the Related Art Conventionally, as a method for performing a voice conference between remote locations, a method of installing a sound emitting and collecting device at each point where a voice conference is performed, connecting these devices through a network, and communicating a voice signal is often used. In many sound emitting and collecting apparatuses, a speaker that emits sound on the partner apparatus side and a microphone that collects sound on the own apparatus side are installed in one casing at the same time.

For example, the audio conference apparatus (sound emitting and collecting apparatus) of Patent Document 1 emits an audio signal input via a network from a speaker arranged on the top surface, and passes through a plurality of different directions arranged on the side surface. Sound is collected by each microphone in the front direction, and the collected sound signal is transmitted to the outside via a network.
JP-A-8-298696

  However, in the apparatus of Patent Document 1, since the microphone and the speaker are close to each other, the sound collected signal of each microphone includes a lot of wraparound sound from the speaker. For this reason, the S / N ratio of the collected signal with respect to the utterance sound of the utterer of the own device is lowered, and the utterance sound of the utterer cannot be clearly collected and output.

  Conventionally, echo cancellation processing exists as a method for removing such wraparound sound. In the echo cancellation process, an adaptive filter and a post processor as a subtractor are used to set a pseudo-regression sound signal for the sneak sound of the input sound signal, that is, the sound emission signal, by the adaptive filter, and from the collected sound signal by the post processor. By subtracting the pseudo-regression sound signal, the wraparound sound is removed.

However, the wraparound audio signal usually has frequency characteristics as shown in FIG.
FIG. 6 is a diagram illustrating frequency characteristics of a general wraparound audio signal. Although this frequency characteristic varies depending on the specifications of the apparatus, the sound emission / collection apparatus configured as described later in this embodiment generally has the frequency characteristic shown in FIG.
As shown in FIG. 6, the wraparound audio signal has a low signal level in the high range exceeding 1000 Hz with respect to the signal level in the low range of about 300 Hz. When echo cancellation processing is performed on a wraparound audio signal having such characteristics with the calculation bit set based on the low-frequency component, the high-frequency component represents the signal level during actual calculation because the signal level is low. The number of bits is reduced, and the calculation accuracy is remarkably lowered with respect to the low frequency range. For this reason, even if low-frequency component echo cancellation can be performed with high accuracy, high-frequency component echo cancellation cannot be performed with high accuracy.

  On the other hand, if the number of calculation bits is increased in order to improve the calculation accuracy for the high frequency component, the calculation processing load is further increased with respect to the echo cancellation process that originally has a larger calculation amount than other processes. For this reason, a higher-performance arithmetic processor (DSP) is required, and the sound emitting and collecting apparatus itself increases in cost.

  Therefore, the object of the present invention is to eliminate the influence of the sneak sound not only on the low frequency but also on the high frequency without requiring a high-performance arithmetic processor or a complicated configuration, and thereby the voice from the speaker. Is to provide a sound emission and collection device that can collect and output sound with a high S / N ratio.

  The sound emission and collection device according to the present invention includes a sound emission means for emitting an input sound signal, a sound collection means for collecting and generating a sound collection signal, and a sound collection means to a sound collection means for the sound collection signal. A wraparound characteristic correction filter that performs a first filtering process that makes the frequency characteristics of the wrapping sound substantially uniform, an echo cancellation unit that performs an echo cancellation process on the collected sound signal that has been subjected to the first filtering process, and an echo that is echoed by the echo cancellation unit And a characteristic readjustment filter that performs a second filtering process having a reverse characteristic of the first filtering process with respect to the canceled sound collection signal.

  In this configuration, a part of the sound emitted from the sound emitting means wraps around and is collected by the sound collecting means. As described above, the wraparound sound collected by the sound collecting means has a high signal level in the low range and a low signal level in the high range.

  The wraparound characteristic correction filter performs a first filtering process that raises the signal level of the high frequency component of the wraparound audio to the signal level of the low frequency component. That is, the signal level of the high frequency component is raised relatively with the low frequency component as a reference, and the frequency characteristics of the signal level are made substantially uniform. At this time, the frequency characteristic of the collected sound signal including the wraparound sound changes according to the first filtering process.

  The echo canceling unit performs echo cancellation processing of the collected sound signal corrected with uniform frequency characteristics for the wraparound sound. At this time, since the signal level of the high frequency component is corrected to be high together with the low frequency component, the high frequency component is expressed by a larger number of effective bits than before the correction (first filtering process). Here, the effective number of bits represents at least the number of bits necessary to represent the signal level of the corresponding frequency. If the signal level is low, the effective number of bits decreases, and if the signal level is high, the effective number of bits increases. . Thus, the increase in the number of effective bits increases the resolution of the signal level of the high frequency component as well as the low frequency component. Therefore, echo cancellation processing is performed with high accuracy in both the low and high frequencies.

  The characteristic readjustment filter performs a second filtering process that is an inverse characteristic of the first filtering process. That is, a signal with a high low-frequency signal level and a low high-frequency signal level is the same as the frequency characteristic of the original wraparound sound with respect to the collected sound signal corrected to a substantially uniform frequency characteristic by the first filtering process. Reverse correction to Thereby, the collected sound signal from which the wraparound sound after the frequency characteristic is corrected is restored to the original frequency characteristic at the time of sound collection.

  As a result, the wraparound sound is removed from the collected sound signal at the time of sound collection, and a sound collected signal whose frequency characteristics do not change is generated.

  Further, the sound collecting means of the sound emitting and collecting apparatus of the present invention has a strong directivity in different directions using a microphone array in which a plurality of microphones are installed in a predetermined arrangement and an audio signal collected by the plurality of microphones. Sound collection beam signal generation means for generating a plurality of sound collection beam signals and outputting the sound collection beam signals as sound collection signals. Furthermore, the wraparound characteristic correction filter and characteristic readjustment filter of the sound emission and collection device of the present invention have a filtering characteristic corresponding to the selected sound collection directivity from a plurality of filtering characteristics set in advance for each sound collection beam signal. It is characterized by selecting and performing a filtering process.

  In this configuration, a plurality of sound collection beam signals are generated using the sound collected by each microphone of the microphone array of the sound collection means. Since each sound collecting beam signal (sound collecting signal) has a different sound collecting directivity, the frequency characteristics of the wraparound sound are also different. The wraparound characteristic correction filter and the characteristic readjustment filter select a filtering characteristic for each sound collection directivity and perform a filtering process, thereby performing a filtering process adapted to the selected sound collection directivity. Thereby, the wraparound sound is removed with higher accuracy according to the sound collection directivity.

  The sound emission means of the sound emission and collection device according to the present invention realizes a plurality of sound emission characteristics different from each other depending on a speaker array in which a plurality of speakers are installed in a predetermined arrangement and a sound emitted by the plurality of speakers. And a sound emission control means for generating sound emission signals for a plurality of speakers. Furthermore, the wraparound characteristic correction filter and characteristic readjustment filter of the sound emission and collection device of the present invention select a filtering characteristic corresponding to the selected sound emission characteristic from a plurality of filtering characteristics set in advance for each sound emission characteristic. And filtering processing.

  In this configuration, a plurality of sound emission characteristics are realized using sound emitted from each speaker of the speaker array of the sound emission means. Each sound emission characteristic has a different frequency characteristic of the wraparound sound. The wraparound characteristic correction filter and the characteristic readjustment filter select the filtering characteristic for each sound emission characteristic and perform the filtering process, whereby the filtering process adapted to the selected sound emission characteristic is performed. Thereby, the wraparound sound is removed with higher accuracy in accordance with the sound emission characteristics.

  The sound emission and collection device of the present invention includes a microphone array in which a plurality of microphones are installed in a predetermined arrangement, and a plurality of microphones having strong directivities in different directions using sound signals collected by the plurality of microphones. A sound collection means comprising a sound collection beam signal generation means for generating a sound collection beam signal and outputting the sound collection beam signal as a sound collection signal; a speaker array comprising a plurality of speakers arranged in a predetermined arrangement; Sound emission means including sound emission control means for generating sound emission signals for the plurality of speakers so that a plurality of sound emission characteristics different from each other are realized by sounds emitted by the speakers. Further, the sneak characteristic correction filter and the characteristic readjustment filter of the sound emission and collection device perform filtering according to a selected combination from a plurality of filtering characteristics set in advance for each combination of the sound collection beam signal and the sound emission characteristic. It is characterized by selecting characteristics and performing filtering processing.

  In this configuration, by performing filtering processing by selecting a filtering characteristic for each combination of sound collection directivity and sound emission characteristic, a filtering process adapted to the selected combination is performed. As a result, depending on the combination of sound collection directivity and sound emission characteristics, the wraparound sound is removed with higher accuracy.

  In addition, the sneak characteristic correction filter and characteristic readjustment filter of the sound emission and collection device of the present invention are characterized in that filtering processing is performed on the echo cancellation means by double-precision bit calculation.

  In this configuration, a high-frequency signal level is corrected and restored with higher accuracy by performing bit operations with double accuracy during the filtering process. Then, while increasing the bit calculation amount of the filtering process, the wraparound sound is removed with higher accuracy without increasing the bit calculation amount of the echo cancellation process. At this time, since the echo cancellation processing has a higher load than the filtering processing, highly accurate wraparound speech is removed without significantly increasing the processing load.

  According to this invention, regardless of the frequency characteristics of the wraparound sound, the wraparound sound is reliably and effectively removed, and the sound from the sound source direction of the speaker or the like is collected and output with a high S / N ratio. Can do.

A sound emitting and collecting apparatus according to an embodiment of the present invention will be described with reference to the drawings.
FIG. 1A is a plan view showing the microphone and speaker arrangement of the sound emitting and collecting apparatus 1 according to the present embodiment, and FIG. 1B is formed by the sound emitting and collecting apparatus 1 shown in FIG. It is a figure which shows a sound collection beam area | region.

  FIG. 2 is a functional block diagram of the sound emitting and collecting apparatus 1 of the present embodiment.

  The sound emission and collection device 1 of the present embodiment includes a housing 101 provided with a plurality of speakers SP1 to SP3, a plurality of microphones MIC11 to MIC17, MIC21 to MIC27, and a functional unit shown in FIG.

  The casing 101 has a substantially rectangular parallelepiped shape that is long in one direction, and has a predetermined height that separates the lower surface of the casing 101 from the installation surface at a predetermined interval at both ends of the long side (surface) of the casing 101. Legs (not shown) are installed. In the following description, of the four side surfaces of the housing 101, a long surface is referred to as a long surface, and a short surface is referred to as a short surface.

  On the lower surface of the housing 101, non-directional single speakers SP1 to SP3 having the same shape are installed. These single speakers SP1 to SP3 are installed in a straight line at regular intervals along the long direction, and the straight line connecting the centers of the single speakers SP1 to SP3 is along the long surface of the casing 101. It is installed such that the horizontal position coincides with the central axis 100 connecting the centers of the short surfaces. That is, a straight line connecting the centers of the speakers SP1 to SP3 is arranged on a vertical reference plane including the central axis 100. As described above, the speaker array SPA 10 is configured by arranging the single speakers SP1 to SP3 in an array. In such a state, when sound is emitted from the individual speakers SP1 to SP3 of the speaker array SPA10, the emitted sound is equally transmitted to the two long surfaces. At this time, the sound emission propagating to two opposing long surfaces proceeds in mutually symmetric directions perpendicular to the reference surface.

  On one long surface of the casing 101, microphones MIC11 to MIC17 having the same specifications are installed. These microphones MIC11 to MIC17 are installed in a straight line at regular intervals along the longitudinal direction, thereby forming a microphone array MA10. In addition, microphones MIC21 to MIC27 having the same specifications are also installed on the other long surface of the casing 101. These microphones MIC21 to MIC27 are also installed in a straight line at regular intervals along the lengthwise direction, thereby forming a microphone array MA20. The microphone array MA10 and the microphone array MA20 are arranged so that the vertical positions of the arrangement axes thereof coincide with each other. The microphones MIC11 to MIC17 of the microphone array MA10 and the microphones MIC21 to MIC27 of the microphone array MA20 are: Each is arranged at a position symmetrical to the reference plane. Specifically, for example, the microphone MIC11 and the microphone MIC21 are symmetrical with respect to the reference plane, and the microphone MIC17 and the microphone MIC27 are similarly symmetrical.

  In the present embodiment, the speaker array SPA10 has three speakers and the microphone arrays MA10 and MA20 each have seven microphones. However, the present invention is not limited to this, and the number of speakers and microphones is not limited thereto. May be set as appropriate. Further, the speaker intervals of the speaker array and the microphone intervals of the microphone array do not have to be constant. For example, they are arranged densely at the center along the longitudinal direction and sparsely arranged toward both ends. Various modes may be used.

  Next, as shown in FIG. 2, the sound emitting and collecting apparatus 1 of this embodiment is functionally composed of a control unit 10, an input / output connector 11, an input / output I / F 12, a sound emitting directivity control unit 13, and a D / A converter 14, sound emission amplifier 15, speaker array SPA10 (speakers SP1 to SP3), microphone arrays MA10 and MA20 (microphones MIC11 to MIC17, MIC21 to MIC27), sound collecting amplifier 16, and A / D converter 17, a sound collection beam generation unit 181, 182, a sound collection beam selection unit 19, and an echo cancellation unit 20.

  The control unit 10 performs operation / process control of the entire device including power supply control and the like, and gives control commands such as arithmetic processing to each unit of the device in response to an operation input command from an operation unit (not shown).

  The input / output I / F 12 converts an input audio signal input from the input / output connector 11 from another sound emitting and collecting device from a data format (protocol) corresponding to the network, and passes through the echo canceling unit 20. To the sound output directivity control unit 13. The input / output I / F 12 converts the output audio signal generated by the echo cancel unit 20 into a data format (protocol) corresponding to the network, and transmits it to the network via the input / output connector 11.

  If the sound emission directivity is not set, the sound emission directivity control unit 13 simultaneously gives a sound emission signal based on the input sound signal to the speakers SP1 to SP3 of the speaker array SPA10. Further, when the sound emission directivity such as the setting of the virtual point sound source is designated by the control unit 10, the sound emission directivity control unit 13 is based on the designated sound emission directivity and each speaker SP1 of the speaker array SPA10. The individual sound emission signal is generated by performing delay processing and amplitude processing specific to .about.SP3 on the input audio signal. The sound emission directivity control unit 13 outputs these individual sound emission signals to the D / A converter 14 installed for each of the speakers SP1 to SP3.

  Each D / A converter 14 converts the individual sound emission signal into an analog format and outputs it to each sound emission amplifier 15, and each sound emission amplifier 15 amplifies the individual sound emission signal and gives it to the speakers SP 1 to SP 3.

  The speakers SP1 to SP3 convert a given sound emission signal or individual sound emission signal into sound and emit the sound outside. Since the speakers SP1 to SP3 are installed on the lower surface of the housing 101, the emitted sound is reflected from the installation surface of the desk on which the sound emitting and collecting apparatus 1 is installed, from the side of the apparatus where the conference person is located. Propagated obliquely upward. Further, a part of the sound emission goes around from the bottom surface of the sound emission and collection device 1 to the side surface on which the microphone arrays MA10 and MA20 are installed.

  The microphones MIC11 to MIC17 and MIC21 to MIC27 of the microphone arrays MA10 and MA20 may be omnidirectional or directional, but are preferably directional, and are external to the sound emitting and collecting apparatus 1. Are collected and electrically converted, and a collected sound signal is output to each sound collecting amplifier 16.

  Further, each of the microphones MIC11 to MIC17 and MIC21 to MIC27 of the microphone arrays MA10 and MA20 collects the wraparound sound of the sound emitted from the speakers SP1 to SP3.

  Each sound collecting amplifier 16 amplifies the collected sound signal and supplies the amplified sound signal to the A / D converter 17, and the A / D converter 17 converts the collected sound signal into a digital signal to the collected sound beam generation units 181 and 182. Output. The collected sound signal from the microphones MIC11 to MIC17 of the microphone array MA10 installed on one long surface is input to the collected sound beam generation unit 181, and the other long length is input to the collected sound beam generation unit 182. The collected sound signals from the microphones MIC21 to MIC27 of the microphone array MA20 installed on the surface are input.

  The collected sound beam generation unit 181 performs predetermined delay processing and the like on the collected sound signals of the microphones MIC11 to MIC17, and generates collected sound beam signals MB11 to MB14. As shown in FIG. 1 (B), the sound collecting beam signals MB11 to MB14 are obtained by collecting areas having different predetermined widths along the long surface on the long surface side where the microphones MIC11 to MIC17 are installed. It is set in the area.

  The collected sound beam generation unit 182 performs predetermined delay processing or the like on the collected sound signals of the microphones MIC21 to MIC27, and generates collected sound beam signals MB21 to MB24. As shown in FIG. 1 (B), the sound collection beam signals MB21 to MB24 are obtained by collecting areas having different predetermined widths along the long surface on the long surface side where the microphones MIC21 to MIC27 are installed. It is set in the area.

  The collected sound beam selection unit 19 performs full-wave rectification, BPF in the speaker sound frequency band, and peak detection on the input sound collected beam signals MB11 to MB14 and MB21 to MB24, and mainly uses the speaker sound. The collected sound collecting beam signal is selected, and the collected sound beam signal MB (corresponding to the “sound collecting signal” of the present invention) is output to the echo canceling unit 20.

  In addition, the sound collection beam selection unit 19 gives the sound collection directivity information corresponding to the selected sound collection beam signal MB to the control unit 10.

  The control unit 10 combines the sound emission directivity information and the sound collection directivity information so as to synchronize according to the delay of the wraparound sound, and gives the combined information to the echo cancellation unit 20.

FIG. 3 is a block diagram showing the main configuration of the echo canceling unit 20 of the present embodiment.
The echo cancellation unit 20 includes a first filter 203, an echo cancellation circuit 200, a second filter 204, and a filter characteristic storage unit 205 in order from the input side of the collected sound beam signal. The echo cancellation circuit 200 includes an adaptive filter 201 and a post processor 202. The adaptive filter 201, the first filter 203, and the second filter 204 are configured by digital filters such as FIR filters, and each filter characteristic is set by each filter coefficient of the digital filter.

  The filter characteristic storage unit 205 stores filter characteristics individually set for each combination of sound emission directivity information and sound collection directivity information with respect to the first filter 203 and the second filter 204. . This filter characteristic depends on the structure of the speaker array SPA10 and the microphone arrays MA10 and MA20, the positional relationship between each speaker SP and each microphone MIC, and the installation status. Just set it up. The setting can be obtained, for example, by collecting a sound emitted with one sound emission directivity and performing processing for each sound collection directivity for all sound emission directivities.

  FIG. 4 is a graph showing the difference in the frequency characteristics of the wraparound sound depending on the combination of the sound emission directivity and the sound collection directivity, and the graphs (A), (B), and (C) are the individual sound emission directivities. In each graph, each characteristic curve indicates a different sound collection directivity Ds11 to Ds14 (sound collection beam signals MB11 to MB14).

  As shown in FIG. 4, in any combination of sound emission directivity and sound collection directivity, a difference in strength occurs between the signal levels of each frequency component. The filter characteristics of the first filter 203 are set such that the signal level is substantially the same in all frequency bands by suppressing the difference in signal level for each frequency component. That is, since the signal level in the low band is high and the signal level in the high band is low, the signal level in the high band is raised to the signal level in the low band. More specifically, the entire frequency band is divided into partial frequency areas each having a predetermined frequency width, and the highest signal level in the low band is used as a reference. Then, the level shift amount of the signal level in each partial frequency region is set so that the signal level substantially matches the reference signal level. At this time, the number of bits higher than the number of calculation bits used in the echo cancellation circuit 200 is used for setting and storing the level shift amount.

  As shown in each of FIGS. 4A to 4C, the frequency characteristics are different for each combination of the sound emission directivity and the sound collection directivity. Therefore, the filter characteristics are set according to the level shift amount. This is performed for each combination of sound emission directivity and sound collection directivity.

  On the other hand, the filter characteristic of the second filter 204 is set so as to cancel the filter characteristic of the first filter 203. That is, the filter characteristic of the second filter 204 is set so that the frequency characteristic of the signal level corrected by the first filter 203 is returned to the original frequency characteristic.

  FIG. 5 is a conceptual diagram showing the contents stored in the filter characteristic storage unit 205. In addition, although this description shows the case where there are three types of sound emission directivities and eight types of sound collection directivities, the number of sound emission directivities and the number of sound collection directivities are not limited to this, and are appropriate. You only have to set it. The filter characteristic storage unit 205 stores a filter characteristic for each combination of individual sound emission directivity and sound collection directivity.

  As illustrated in FIG. 5, the filter characteristic storage unit 205 stores filter characteristics for each combination of the sound emission directivities Dv1, Dv2, and Dv3 and the sound collection directivities Ds11 to Ds14 and Ds21 to Ds24. For example, if the sound emission directivity Dv1 and the sound collection directivity Ds11 are combined, the first filter characteristic Fc111 and the second filter characteristic Fr111 are stored. Similarly, if the sound emission directivity Dv3 and the sound collection directivity Ds24 are combined, the first filter characteristic Fc324 and the second filter characteristic Fr324 are stored. That is, in the example shown in FIG. 5, 3 × 8 = 24 kinds of first filter characteristics Fc and second filter characteristics Fr are stored.

  When the echo cancel unit 20 receives the combination information of the sound emission directivity information and the sound collection directivity information from the control unit 10, the echo cancel unit 20 transmits the corresponding first filter characteristic Fc and second filter characteristic Fr from the filter characteristic storage unit 205. read out. The echo cancellation unit 20 sets each filter coefficient of the first filter 203 based on the read first filter characteristic Fc, and sets each filter coefficient of the second filter 204 based on the read second filter characteristic Fr.

  The first filter 203 performs a filtering process based on the first filter characteristic Fc on the input sound collection beam signal MB, and outputs the filtered signal to the post processor 202 of the echo cancellation circuit 200. That is, the signal level of the high frequency range of the collected sound beam signal MB is increased so that the frequency characteristic due to the wraparound is substantially uniform from the low frequency range to the high frequency range. As a result, the change in the frequency spectrum due to the wraparound is canceled and the signal level is different, but this is equivalent to that the wraparound sound having the same frequency spectrum as that of the input sound signal input to the sound output directivity control unit 13 is captured. In addition, since the signal level is uniform in both the low and high frequencies, the number of effective bits representing the signal level at each frequency increases, and the resolution increases in the entire frequency band.

  For this level shift calculation, a higher number of bits than the number of calculation bits used in the echo cancellation circuit 200 is used. For example, if the number of operation bits of the echo cancel circuit 200 is 16, the number of bits for level shift operation is set to 32 bits. Further, floating point arithmetic may be used for the level shift. As a result, a large amount of bits can be allocated to the high region where the signal level is originally low, so that the signal level of the corrected sound collection beam signal MB can be calculated with higher accuracy. That is, it is possible to suppress quantization errors during signal correction.

  The adaptive filter 201 of the echo cancellation unit 20 generates a pseudo regression sound signal based on the sound collection directivity of the selected sound collection beam signal MB with respect to the input sound signal. At this time, as an initial condition, the adaptive filter 201 assumes that the sound circulates equally in all frequency bands from the low range to the high range regardless of the sound emission directivity and the sound collection directivity. Start generation. Thereby, the generation start conditions of the pseudo regression sound signal are simplified (unified), and an increase in calculation load can be prevented. The post processor 202 functions as a subtracter, subtracts the pseudo regression sound signal from the collected sound beam signal MB corrected by the first filter 203, and outputs it to the second filter 204. Here, the collected sound beam signal input to the post processor 202 includes the wraparound sound whose signal level is substantially uniformly corrected in the entire frequency band as described above, so the input sound signal and the corrected wraparound sound are included. The frequency spectrum of the signal substantially matches. Therefore, the frequency spectra of the pseudo-regression sound signal based on the input sound signal and the wraparound sound signal included in the collected sound beam signal MB also substantially match. In addition, the effective number of bits in the entire frequency band increases, and the resolution increases. Thereby, the wraparound sound can be reliably removed with high accuracy from the collected sound beam signal MB.

  The second filter 204 performs a filtering process based on the second filter characteristic Fr on the collected sound beam signal from which the wraparound sound is removed, and outputs it to the input / output I / F 12 as an output sound signal. The second filter characteristic Fr lowers the high-frequency signal level raised by the first filter characteristic Fc so as to conform to the frequency characteristic of the original wraparound sound. As a result, the frequency characteristics of the speech signal other than the wraparound speech signal, that is, the utterance sound of the speaker, whose high frequency is raised together with the wraparound speech by the first filter 203, are readjusted to the state before the correction by the first filter 203. can do. That is, it is possible to obtain an output audio signal according to the collected sound beam signal based on the collected raw frequency spectrum.

  Also for this readjustment calculation, the number of bits higher than the number of calculation bits used in the echo cancellation circuit 200 is used. For example, if the number of operation bits of the echo cancellation circuit 200 is 16 bits, the number of readjustment operations is set to 32 bits. In addition, a floating point operation may be used for the readjustment operation. Thereby, the signal level of the readjusted sound collecting beam signal MB can be calculated with high accuracy. That is, it is possible to suppress a quantization error during signal readjustment.

  In addition, by using double precision for the first and second filter operations having a relatively low calculation processing load as compared to the calculation accuracy of the echo cancellation processing having a high calculation processing load, the echo cancellation processing is originally complicated and has a high load. Without further increasing the computational load, it is possible to obtain a sound collection beam signal (output sound signal) from which the wraparound sound is removed with high accuracy.

  As described above, by using the configuration and processing of this embodiment, the wraparound speech is removed with high accuracy, and only the necessary sound such as the utterance sound of the speaker is acquired and output with a high S / N ratio. A sound collection device can be realized.

  In the above description, the example in which both the first filter characteristic Fc and the second filter characteristic Fr are stored in the filter characteristic storage unit 205 has been described. However, since the second filter characteristic Fr is a reverse correction characteristic with respect to the first filter characteristic Fc, only the first filter characteristic Fc is stored, and the second filter characteristic Fr is calculated from the selected first filter characteristic Fc. Thus, the second filter 204 may be set.

  In the above description, the sound emitting and collecting apparatus including both the microphone array having the sound collecting directivity and the speaker array having the sound emitting directivity has been described. Even when a combination with a speaker array having sound emission directivity or a combination of a microphone array with sound collection directivity and a speaker and speaker array without sound emission directivity is used, the above-described configuration can be applied. it can.

It is the top view which shows the microphone of the sound emission and collection apparatus which concerns on this embodiment, and speaker arrangement | positioning, and the figure which shows the sound collection beam area | region formed with a sound emission and collection apparatus. It is a functional block diagram of the sound emission and collection device of this embodiment. It is a block diagram which shows the main structures of the echo cancellation part 20 of this embodiment. It is a graph which shows the difference in the frequency characteristic of the wraparound sound by the combination of sound emission directivity and sound collection directivity. FIG. 4 is a conceptual diagram showing storage contents of a filter characteristic storage unit 205. It is a figure which shows the rough frequency characteristic of a wraparound sound.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1- Sound emission / collection apparatus, 101- Housing | casing, 11- Input / output connector, 12- Input / output I / F, 13- Sound emission directivity control part, 14-D / A converter, 15- Sound emission amplifier, 16 -Amplifier for sound collection, 17-A / D converter, 181,182 -Sound collecting beam generation unit, 19 -Sound collecting beam selection unit, 20 -Echo canceling unit, 201 -Adaptive filter, 202 -Post processor, 2203, 204-filter, SP1-SP3-speaker, SPA10-speaker array, MIC11-MIC17, MIC21-MIC27-microphone, MA10, MA20-microphone array

Claims (5)

Sound emission means for emitting an input audio signal;
Sound collection means for collecting sound and generating a sound collection signal;
A wraparound characteristic correction filter that performs a first filtering process for making the frequency characteristic of the sound that wraps around from the sound emitting means to the sound collecting means substantially uniform with respect to the sound collected signal;
Echo cancellation means for performing echo cancellation processing on the collected sound signal subjected to the first filtering processing;
A characteristic readjustment filter that performs a second filtering process consisting of the reverse characteristics of the first filtering process on the collected sound signal that has been echo canceled by the echo canceling means;
A sound emission and collection device. The sound collecting means uses a microphone array in which a plurality of microphones are installed in a predetermined arrangement, and a plurality of sound collecting beams having strong directivities in different directions using sound signals collected by the plurality of microphones. A sound collecting beam signal generating means for generating a signal and outputting the sound collecting beam signal as the sound collecting signal;
The wraparound characteristic correction filter and the characteristic readjustment filter perform a filtering process by selecting a filtering characteristic corresponding to a selected collected sound beam signal from a plurality of filtering characteristics set in advance for each collected sound beam signal. Item 2. The sound emission and collection device according to Item 1. The sound emitting means includes a speaker array in which a plurality of speakers are arranged in a predetermined arrangement, and a plurality of sound emission characteristics that are different from each other by sounds emitted from the plurality of speakers. A sound emission control means for generating a sound emission signal for the speaker;
2. The wraparound characteristic correction filter and the characteristic readjustment filter perform a filtering process by selecting a filtering characteristic corresponding to the selected sound emission characteristic from a plurality of filtering characteristics set in advance for each sound emission characteristic. The sound emission and collection device described in 1. The sound emitting means includes a speaker array in which a plurality of speakers are arranged in a predetermined arrangement, and a plurality of sound emission characteristics that are different from each other by sounds emitted from the plurality of speakers. A sound emission control means for generating a sound emission signal for the speaker;
The wraparound characteristic correction filter and the characteristic readjustment filter perform filtering by selecting a filtering characteristic corresponding to the selected combination from a plurality of filtering characteristics set in advance for each combination of the collected sound beam signal and the sound emission characteristic. The sound emission and collection device according to claim 2 which performs processing.   The sound emission and collection device according to any one of claims 1 to 4, wherein the wraparound characteristic correction filter and the characteristic readjustment filter perform a filtering process by double-precision bit calculation on the echo cancellation unit.

Images