JP7697497B2 - Sound collection device, sound collection program, and sound collection method - Google Patents

Description

This invention relates to a sound collection device, a sound collection program, and a sound collection method, and can be applied to area sound collection processing that suppresses sounds outside a specific area and emphasizes only sounds from that specific area, for example, in order to recognize only sounds emitted from that specific area.

When using a speech recognition system in a noisy environment, ambient noise that mixes in with the desired target sound can be a nuisance, reducing the speech recognition rate of the recorded voice.

Conventionally, in an environment where multiple sound sources exist, beamforming using a microphone array is a technology that collects only sounds from a specific direction and obtains the required target sound while avoiding the inclusion of unwanted sounds. Beamforming is a technology that forms directivity by utilizing the time difference between signals arriving at each microphone (see Non-Patent Document 1). A microphone array that performs beamforming is also called a beamformer.

Beamformers, on the other hand, pick up all sounds from a specific direction, so if you want to pick up only a specific area (hereafter also referred to as the "target area"), they will also pick up sounds outside the target area in the same direction as the beamformer. Therefore, Patent Document 1 proposes a method (area sound collection) that uses multiple beamformers, directs their directivity toward the target area from different directions, and crosses the directivity at the target area to pick up the target sound.

JP 2013-183358 A

Futoshi Asano, "Acoustic Technology Series 16: Array Signal Processing of Sound - Localization, Tracking and Separation of Sound Sources", edited by the Acoustical Society of Japan, Corona Publishing, published on February 25, 2011

When collecting human voices, it is desirable for the beamformer to have sharp directivity in order to form a target area. For example, when collecting human voices, it is desirable for the beamformer's beam width to be approximately 60 degrees or less.

In both additive and subtractive beamformers, a large number of microphones are required to form sharp directivity. For example, the Minimum Variance Distortionless Response (MVDR) method requires one more microphone than the number of noise sources to be suppressed. Although area recording with such a large number of microphones can be experimentally achieved in a laboratory, it is not suitable for commercial products in terms of component costs and installation space.

On the other hand, the technology described in Patent Document 1 uses spectral subtraction in the frequency domain (a method in which amplitude spectra are subtracted from each other, but if the subtraction result is negative, it is replaced with zero or a small number) to achieve sharp directivity with a small number of microphones. A null former that forms a blind spot in the direction of the target area is created, and sharp directivity is formed in the direction of the target area by spectrally subtracting the null former amplitude spectrum from the microphone input amplitude spectrum. Using this method, sharp directivity can be formed with just two microphones, but it causes distortions such as musical noise and missing area sound components.

Therefore, with conventional technology, there were issues such as the need for multiple microphones to achieve area sound collection, which resulted in large equipment size, and the use of spectral subtraction with a small number of microphones resulted in distortion of the extracted area sound.

In view of the above problems, there is a need for a sound collection device, a sound collection program, and a sound collection method that suppress distortion of sounds collected when performing area sound collection processing to collect sounds from a target area.

The first invention has a null-former processing means for calculating a plurality of null-formers forming a blind spot outside a target area for each of a plurality of microphone arrays that can be configured by a combination of the microphones of the microphone group based on an acoustic signal supplied from a microphone group consisting of three or more microphones, and acquiring a null-former output sound, and an area sound extraction means for acquiring an area sound obtained by extracting a sound of the target area based on the null-former output sound of each of the null formers, and the area sound extraction means calculates a plurality of the null-former output sounds for each of the microphone arrays for each frequency. the null former selection unit selecting the null former output sound having the smallest output from among the force sounds as the beamformer output sound of the microphone array, and the beamformer selection unit selecting the beamformer output sound having the smallest output from among the plurality of beamformer output sounds as an area sound , wherein the null former processing means acquires the null former output sounds in equal numbers from a first azimuth angle range ranging from -π/2 to 0 and excluding the target area direction, and a second azimuth angle range ranging from π/2 to 0 and excluding the target area direction, for each of the microphone arrays .

A sound collection program according to a second aspect of the present invention causes a computer to function as a null-former processing means for calculating, based on an acoustic signal supplied from a microphone group consisting of three or more microphones, a plurality of null formers forming blind spots outside a target area for each of a plurality of microphone arrays that can be configured by a combination of the microphones of the microphone group, and acquiring a null-former output sound, and an area sound extraction means for acquiring an area sound obtained by extracting a sound of the target area based on the null-former output sound of each of the null formers, and the area sound extraction means calculates a plurality of null formers for each frequency for each of the microphone arrays. The system includes a null former selection unit that selects the null former output sound with the smallest output from the null former output sounds and sets it as a beamformer output sound of the microphone array, and a beamformer selection unit that selects the beamformer output sound with the smallest output from a plurality of the beamformer output sounds and sets it as an area sound, and the null former processing means acquires the null former output sounds in equal numbers from each of a first azimuth angle range ranging from -π/2 to 0 and excluding the target area direction, and a second azimuth angle range ranging from π/2 to 0 and excluding the target area direction, for each of the microphone arrays .

The third aspect of the present invention is a sound collection method performed by a sound collection device, the sound collection device comprising a null former processing means and an area sound extraction means, the area sound extraction means having a null former selection unit and a beamformer selection unit, the null former processing means calculating a plurality of null formers forming blind spots outside a target area for each of a plurality of microphone arrays that can be configured by a combination of the microphones of the microphone group based on an acoustic signal supplied from a microphone group consisting of three or more microphones, and acquiring a null former output sound, and the area sound extraction means acquiring an area sound obtained by extracting a sound of the target area based on the null former output sound of each of the null formers. The null former selection unit selects, for each of the microphone arrays, the null former output sound with the smallest output from the multiple null former output sounds for each frequency to set it as the beamformer output sound of the microphone array, the beamformer selection unit selects the beamformer output sound with the smallest output from the multiple beamformer output sounds to set it as an area sound , and the null former processing means acquires the same number of null former output sounds from each of a first azimuth angle range ranging from -π/2 to 0 and excluding the target area direction, and a second azimuth angle range ranging from π/2 to 0 and excluding the target area direction, for each of the microphone arrays .

The present invention provides a sound collection device, a sound collection program, and a sound collection method that suppress distortion of sounds collected when performing area sound collection processing to collect sounds in a target area.

1 is a block diagram showing a functional configuration of a sound collection device according to a first embodiment. FIG. 1 is a diagram (part 1) illustrating an example of the positional relationship between a microphone array and a target area according to the first embodiment. FIG. 2 is a diagram (part 2) illustrating an example of the positional relationship between the microphone array and the target area according to the first embodiment. 1 is a block diagram showing an example of a hardware configuration of a sound collection device according to a first embodiment. 1A to 1C are diagrams showing examples of directions of null formers applied to an arbitrary microphone array according to the first embodiment. FIG. 1 is a table showing the characteristics of sounds selected by the minimum selection of the beamformer selection unit according to the first embodiment; FIG. 11 is a block diagram showing a functional configuration of a sound collection device according to a second embodiment. FIG. 11 is a diagram (part 1) showing an example of the positional relationship between a microphone array and a target area according to the second embodiment; FIG. 13 is a diagram (part 2) illustrating an example of the positional relationship between a microphone array and a target area according to the second embodiment.

(A) First Embodiment Hereinafter, a first embodiment of a sound collection device, a sound collection program, and a sound collection method according to the present invention will be described in detail with reference to the drawings.

(A-1) Configuration of the First Embodiment FIG. 1 is a block diagram showing the functional configuration of a sound collection device 10 according to the first embodiment.

The sound collection device 10 suppresses sounds emitted outside the target area and collects sounds emitted within the target area based on the acoustic signals captured by the two microphone arrays MA (MA1, MA2).

In this embodiment, the microphone arrays MA1 and MA2 are each assumed to be microphone arrays equipped with two microphones. The microphone array MA1 is equipped with two microphones M (M1, M2), and the microphone array MA2 is also equipped with two microphones M (M3, M4). In other words, in the first embodiment, the two microphone arrays MA1 and MA2 are configured using a microphone group consisting of four microphones M1 to M4.

In this embodiment, the sound collection device 10 is described as a device that supports two configurations with a microphone array having two microphones, but it may be configured to support three or more microphone arrays, and each microphone array may be configured to have three or more microphones. However, from the standpoint of parts cost and installation space, it is desirable to keep the number of microphone arrays/microphones that the sound collection device 10 of this embodiment supports as small as possible, so the minimum configuration shown in Figure 1 (two 2ch microphone array configuration) is most suitable.

Figure 2 shows an example of the positional relationship between two microphone arrays (microphone arrays MA1 and MA2) and the target area TA.

In FIG. 2, the area where the front direction of the microphone array MA1 and the front direction of the microphone array MA2 intersect is illustrated as the target area TA. In FIG. 2, the target area TA is hatched (diagonal lines). In this embodiment, as shown in FIG. 2, the number of microphones constituting each microphone array is two, and the number of microphone arrays is two, but the number of microphones and microphone arrays may be increased as long as each is two or more. For example, as shown in FIG. 3, a configuration in which three microphone arrays each having four microphones are arranged (three microphone arrays are arranged from a microphone group consisting of 12 microphones) may be used. In FIG. 3, the microphone array MA1 having microphones M1 to M4, the microphone array MA2 having microphones M5 to M8, and the microphone array MA3 having microphones M9 to M12 are arranged around the target area TA. In FIG. 3, the area where the front directions of the microphone arrays MA1 to MA3 intersect is illustrated as the target area TA, and is hatched (diagonal lines).

In Figure 3, the destination area TA is hatched (diagonal lines).

Next, the internal configuration of the sound collection device 10 will be described with reference to FIG.

The sound collection device 10 has a null former processing unit 101, a null former selection unit 102, and a beam former selection unit 103.

The null former processing unit 101 forms multiple null formers, each with a blind spot facing outside the target area (in a direction not including the target area), based on the acoustic signals captured by the two microphone arrays MA1 and MA2, and obtains multiple null former output sounds. In the first embodiment, the null former processing unit 101 is equipped with a means (hereinafter referred to as a "signal input means") for converting the acoustic signals captured by the microphone arrays MA1 and MA2 from analog signals to digital signals and further converting them into the frequency domain.

The null former selection unit 102 obtains a sound (hereinafter referred to as "beamformer output sound") with directivity formed in the direction of the target area for each microphone array based on the multiple null former output sounds calculated by the null former processing unit 101.

The beamformer selection unit 103 obtains the result of extracting the target area sound (hereinafter referred to as "area sound") based on the beamformer output sound calculated for each microphone array by the null former selection unit 102.

As described above, in the first embodiment, the null former selection unit 102 and the beam former selection unit 103 constitute an area sound extraction means that performs area sound collection processing using a null former.

Figure 4 is a block diagram showing an example of the hardware configuration of the sound collection device 10.

Figure 4 shows an example of the hardware configuration when the sound collection device 10 is configured using software (computer).

The sound collection device 10 shown in FIG. 4 has, as a hardware component, a computer 300 on which a program (including the sound collection program of the embodiment) is installed. The computer 300 may be a computer dedicated to the sound collection program, or may be configured to be shared with programs of other functions.

The computer 300 shown in FIG. 4 has a processor 301, a primary storage unit 302, and a secondary storage unit 303. The primary storage unit 302 is a storage unit that functions as a working memory (work memory) for the processor 301, and can be, for example, a high-speed memory such as a dynamic random access memory (DRAM). The secondary storage unit 303 is a storage unit that records various data such as an operating system (OS) and program data (including data of the sound collection program according to the embodiment), and can be, for example, a non-volatile memory such as a FLASH (registered trademark) memory, HDD, or SSD. In the computer 300 of this embodiment, when the processor 301 starts up, the OS and programs (including the sound collection program according to the embodiment) recorded in the secondary storage unit 303 are read, deployed on the primary storage unit 302, and executed.

The specific configuration of the computer 300 is not limited to the configuration in FIG. 4, and various configurations can be applied. For example, if the primary storage unit 302 is a non-volatile memory (e.g., a flash memory, etc.), the secondary storage unit 303 may be excluded.

(A-2) Operation of the First Embodiment Next, the operation of the sound collection device 10 of the first embodiment having the above-mentioned configuration (sound collection method according to the embodiment) will be described.

In the null former processing unit 101, the signal input means converts the acoustic signals captured by the microphone arrays MA1 and MA2 (microphones M1 to M4) from analog signals to digital signals, performs arbitrary frequency conversion, and obtains the resulting frequency spectrum (hereinafter referred to as the "input spectrum"). The frequency conversion method for obtaining the input spectrum by the signal input means is most preferably, but is not limited to, the fast Fourier transform, and discrete Gabor transform, wavelet transform, filter bank, etc. may also be used.

The null former processing unit 101 calculates, for each microphone array, multiple null formers that form blind spots in specific directions other than the target area based on the supplied input spectrum, and supplies the obtained null former spectrum (hereinafter referred to as "null former output sound") to the null former selection unit 102.

Next, we will explain the details of the processing of the null formatter processing unit 101.

Figure 5 shows an example of the direction of the null former in an arbitrary microphone array MAa.

In the following, the "a" in "MAa" etc. indicates the microphone array number (identifier; ID). Here, the microphone array MA1 is numbered 1, and the microphone array MA2 is numbered 2.

In other words, the microphone array MAa shown in FIG. 5 indicates either the microphone array MA1 or MA2 (any microphone array). FIG. 5 illustrates that the microphone array MAa includes microphones ML and MR. For example, when the microphone array MAa is the microphone array MA1, the microphones ML and MR are microphones M1 and M2, respectively, and when the microphone array MAa is the microphone array MA2, the microphones ML and MR are microphones M3 and M4, respectively. In the image diagram of FIG. 5, the shape of the target area TA is illustrated as a circle for convenience of illustration, but in this embodiment, the actual shape of the target area TA is as shown in FIG. 2.

Here, the front direction (0 degree direction) of the microphone array MAa is defined as the direction that is perpendicular to the line connecting the positions (center positions) of the two microphones ML and MR and that forms the smallest angle with the direction from the center point of the two microphones ML and MR toward the center point (which can be the center of gravity) of the target area TA.

As shown in Figure 2, the target area TA is not a point but has a range, so the apparent area direction from each microphone array also has a range. Therefore, here, the target area direction (range of target area direction) in microphone array MA1 is θ1L to θ1R (rad) (-π/2<θ1L<0<θ1R<π/2), and the target area direction (range of target area direction) in microphone array MA2 is θ2L to θ2R (rad) (-π/2<θ2L<0<θ2R<π/2).

As described above, the specified blind spot direction applied to each null former of the null former processing unit 101 must be outside the target area direction (a direction that does not include the range of the target area direction). Therefore, the nth blind spot direction φan(rad) (n=1,...,N) of the microphone array MAa (a=1,2) must satisfy -π-θaL<φan<θaL or θaR<φan<π-θaL.

Here, since the microphone array MAa consists of two microphones, if a blind spot is formed in φan, a blind spot is also formed in π-φan (=-π-φan). Also, since it is difficult to know in advance from which direction noise will come, it is desirable to determine the specified blind spot direction in advance by design (applying a predesigned fixed value).

From the above, it is preferable to determine the same number of predetermined blind spot directions to be applied to each null former of the null former processing unit 101 from the two ranges of -π/2≦φan<θaL and θaR<φan≦π/2. In the null former processing unit 101, the number of predetermined blind spot directions set for each microphone array (the number of null formers set for each microphone array MAa in the null former processing unit 101) and the combination of blind spot directions are not limited. For example, when the number N of predetermined blind spot directions is 6 (when N=6), it is possible to select φa1=-π/2, φa2=-π/3, φa3=-π/6, φa4=π/6, φa5=π/3, and φa6=π/2, and this selection method is preferable.

Next, we will explain examples of specific processing of each null former in the null former processing unit 101.

When the input spectrum of the microphone array MAa is X _aL (ω) and X _aR (ω), the null former processing unit 101 calculates a plurality of null former spectra Y _an (ω) using equation (1), where ω (omega) is the angular frequency, i is the imaginary unit, d is the microphone spacing, and c is the speed of sound.

Next, we will explain the specific processing of the null format selection unit 102.

The null former selection unit 102 selects the null former spectrum with the smallest amplitude for each microphone array and for each frequency based on the multiple null former spectra supplied, acquires the obtained spectrum (hereinafter referred to as the "beamformer spectrum") as a beamformer output sound, and supplies it to the beamformer selection unit 103. Specifically, the null former selection unit 102 calculates the beamformer spectrum (beamformer output sound) B _a (ω) derived from the microphone array MAa using equations (2) and (3). Here, v(ω) (v: nu) is the index number of the blind angle at which the null former spectrum is the smallest, and argmin_n{} is an operator that outputs n that minimizes the value in {}.

That is, the beamformer spectrum _B1 (ω) becomes sound in the target area direction (range in the target area direction) at the microphone array MA1, and the beamformer spectrum _B2 (ω) becomes sound in the target area direction (range in the target area direction) at the microphone array MA2. The null former selection unit 102 supplies the beamformer spectrum _B1 (ω) of the microphone array MA1 and the beamformer spectrum _B2 (ω) of the microphone array MA2 to the beamformer selection unit 103.

Next, we will explain the specific processing of the beamformer selection unit 103.

The beamformer selection unit 103 acquires a spectrum (hereinafter referred to as an "area sound spectrum") obtained by selecting the beamformer spectrum _B1 (ω) and _B2 (ω) with the smallest amplitude for each frequency from the supplied beamformer spectra B1(ω) and B2(ω) as the area sound. Specifically, the beamformer selection unit 103 calculates the area sound spectrum Z(ω) using equations (4) and (5).

In addition, in Patent Document 1, the area sound spectrum is extracted by performing two spectral subtractions on the amplitude of the beamformer spectrum, but in this embodiment, the same processing can be achieved by using equations (4) and (5).

Figure 6 is a table showing the characteristics of sounds selected by the beamformer selection unit 103 in minimum selection.

In FIG. 6, the range of the first beamformer is the range of the directivity of the beamformer spectrum B ₁ (ω) of the microphone array MA1, and the range of the second beamformer is the range of the directivity of the beamformer spectrum B ₂ (ω) of the microphone array MA2.

As shown in FIG. 6, when sound exists within the range of both beamformers for a certain frequency (when sound exists within the target area), the target area sound component will be selected regardless of which beamformer's sound (amplitude) is selected. Also, as shown in FIG. 6, when sound exists only within the range of one of the beamformers, the beamformer on the side where sound does not exist will be selected by minimum selection. Furthermore, as shown in FIG. 6, when sound exists outside the range of either beamformer, sound outside the target area will not be selected regardless of which beamformer's sound (amplitude) is selected. Therefore, as shown in FIG. 6, the minimum selection process in the beamformer selection unit 103 results in sound being picked up only when sound exists within the range of both beamformers (i.e., when area sound is generated within the target area), and when it is not (i.e., when sound is generated outside the target area), the sound is not picked up.

(A-3) Advantages of the First Embodiment According to the first embodiment, the following advantages can be achieved.

The sound collection device 10 of the first embodiment suppresses sounds coming from directions other than the target area direction, forms a beamformer that collects only the target area direction, and collects the target area sound using the output of this beamformer. This allows the sound collection device 10 of the first embodiment to collect only the area sound without causing distortion such as that caused by spectral subtraction.

(B) Second Embodiment Hereinafter, a second embodiment of the sound collection device, the sound collection program, and the sound collection method according to the present invention will be described in detail with reference to the drawings.

(B-1) Configuration of the second embodiment FIG. 7 is a block diagram showing the functional configuration of a sound collection device 10A according to the second embodiment, and parts that are the same as or correspond to those in FIG. 1 above are given the same or corresponding symbols.

In the first embodiment, there was an explicit constraint that a microphone array must be prepared when placing the microphones (two 2-channel microphone arrays must be prepared in the example of the first embodiment). However, looking at equations (2) to (5), it can be seen that the area sound spectrum can be obtained by selecting the null former spectrum with the smallest output, regardless of which microphone array the components contained in the target area sound ultimately originate from.

The second embodiment of the sound collection device 10A therefore sets up two or more microphone arrays using a microphone group with three or more microphones arranged freely to surround the target area, and extracts the area sound spectrum by selecting the null former spectrum of all the microphone arrays with the smallest output.

Next, the internal configuration of the sound collection device 10A will be described using FIG. 7.

The sound collection device 10A acquires null forms for two or more microphone arrays based on the acoustic signals captured by M (M is 3 or more) microphones M1 to MM, suppresses sounds emitted outside the target area based on each acquired null form, and collects sounds emitted within the target area.

Figures 8 and 9 show an example of the configuration of a microphone array in the second embodiment.

In FIG. 8, three microphones M1 to M3 are arranged around the target area TA (M=3), and two microphone arrays MA1 and MA2 are formed by the three microphones M1 to M3. In FIG. 9, four microphones M1 to M4 are arranged around the target area TA (M=4), and three microphone arrays MA1 to MA3 are formed by the four microphones M1 to M4. In FIG. 8 and FIG. 9, the area where the front directions of the microphone arrays intersect is the target area TA, and is hatched (diagonal lines). The target area TA is hatched (diagonal lines). In the second embodiment, as shown in FIG. 8 and FIG. 9, microphones (microphone group) that are not previously configured as a microphone array are arranged around the target area TA, and the sound collection device 10A is configured to use any combination of microphones as a microphone array. 8 and 9 show an example in which the microphone array is formed by combining the closest microphones around the target area TA, but it is not necessary to form the microphone array with the closest microphones. For example, in FIG. 9, the microphone array may be formed with microphones M1 and M3.

As shown in FIG. 7, the sound collection device 10A has a microphone array selection unit 201, a null former processing unit 101A, and a null former selection unit 202.

The microphone array selection unit 201 selects multiple combinations from M microphones M1 to MM, and configures (functions) each combination (combination of multiple microphones) as a microphone array. The microphones for each microphone array can be freely selected as long as different microphone arrays do not have exactly the same combination of microphones, but they must be selected so that a target area TA can be formed (in other words, so that the common portion of the directivities of multiple beamformers formed later can have a closed shape). Note that combinations in which some microphones are common between different microphone arrays may be used, as shown in Figures 8 and 9.

In the second embodiment, the most suitable configuration is that shown in FIG. 2 (two sets of 2ch microphone arrays), which allows relatively free design of the target area TA while reducing the number of microphones. However, for example, as shown in FIG. 3, three microphone arrays may be configured with 12 microphones (three sets of 4ch microphone arrays).

In the second embodiment, the microphone array selection unit 201 is equipped with a signal input means for converting the acoustic signals captured by the microphones M1 to MM from analog signals to digital signals, and further converting them into the frequency domain.

The null former processing unit 101A forms multiple null formers each facing a blind spot outside the target area based on the acoustic signal captured by the microphone selected by each microphone array, and obtains multiple null former output sounds. In other words, the null former processing unit 101A forms multiple null formers facing a blind spot outside the target area for each microphone array configured by the microphone array selection unit 201, and obtains null former output sounds.

The null former selection unit 202 obtains an area sound based on the multiple null former output sounds calculated by the null former processing unit 101A.

As described above, in the second embodiment, the null former selection unit 202 constitutes an area sound extraction means that performs area sound collection processing using a null former.

(B-2) Operation of the Second Embodiment Next, the operation of the sound collection device 10A of the second embodiment having the above-mentioned configuration (sound collection method according to the embodiment) will be described.

In the microphone array selection unit 201, the signal input means converts the acoustic signals captured by the microphones M1 to MM from analog signals to digital signals, and then performs an arbitrary frequency conversion, and the resulting frequency spectrum is acquired as the input spectrum. The processing of the signal input means can be the same as that in the first embodiment, so a detailed explanation will be omitted.

The microphone array selection unit 201 assigns microphones to a predetermined number of R microphone arrays (R is an integer equal to or greater than 2) in accordance with a predetermined microphone selection list or the like, and supplies the input spectrum supplied from microphones M1 to MM to the null former processing unit 101A for each configured microphone array. Note that the number of microphone arrays set by the microphone array selection unit 201 and the selection process of the microphones associated with each microphone array may be performed based on previously set (designed) parameters or programs, or may be changed dynamically.

The null former processing unit 101A calculates multiple null formers that form blind spots in specific directions other than the target area for each microphone array based on the supplied input spectrum, and supplies the obtained null former spectrum (null former output sound) to the null former selection unit 202.

The detailed operation of the null former processing unit 101A is similar to that of the null former processing unit 101 according to the first embodiment, and therefore a detailed description thereof will be omitted. Here, the multiple null former spectra obtained by the null former processing unit 101A will be denoted as Y _an (ω) as in the first embodiment.

The null former selection unit 202 selects a null former spectrum with a minimum amplitude for each frequency based on the multiple null former spectra supplied, and outputs the obtained area sound spectrum (area sound). Specifically, the null former selection unit 202 calculates the area sound spectrum Z(ω) using equations (6) and (7).

(B-3) Advantages of the Second Embodiment In addition to the advantages of the first embodiment, the second embodiment can achieve the following advantages.

In the sound collection device 10A of the second embodiment, the microphone array selection unit 201 can set a microphone array in any combination from a group of microphones, making it possible to select the microphone arrangement and microphone array configuration more flexibly than in the first embodiment.

(C) Other Embodiments The present invention is not limited to the above-described embodiments, and modified embodiments such as those exemplified below can also be mentioned.

(C-1) In each of the above embodiments, the sound collection devices 10, 10A are shown to have a signal input means (means for converting the acoustic signal captured by each microphone from an analog signal to a digital signal and then converting it to the frequency domain), but the signal input means may be arranged outside the sound collection devices 10, 10A (arranged between each microphone and the sound collection devices 10, 10A).

10, 10A...sound collection device, 101...null former processing unit, 101A...null former processing unit, 102...null former selection unit, 103...beamformer selection unit, 201...microphone array selection unit, 202...null former selection unit, M1 to MM...microphones, MA1 to MA3...microphone array.

Claims (4)

a null-former processing means for calculating a plurality of null-formers that form blind spots outside a target area for each of a plurality of microphone arrays that can be configured by a combination of the microphones of the microphone group based on an acoustic signal supplied from the microphone group that is configured from three or more microphones, and acquiring a null-former output sound;
and an area sound extraction means for acquiring an area sound obtained by extracting a sound of the target area based on the null former output sound of each of the null formers,
The area sound extraction means
a null former selection unit that selects the null former output sound having the smallest output from the plurality of null former output sounds for each frequency of the microphone array as the beamformer output sound of the microphone array;
a beamformer selection unit that selects the beamformer output sound having the smallest output from the plurality of beamformer output sounds as an area sound ,
The null former processing means acquires the null former output sounds from a first azimuth angle range from −π/2 to 0 and excluding the target area direction, and a second azimuth angle range from π/2 to 0 and excluding the target area direction, for each of the microphone arrays, by the same number.
A sound collecting device characterized by the above. a microphone array selection unit that selects a plurality of combinations of microphones from the group of microphones and configures the microphone array for each selected combination;
the null former processing means calculates a plurality of null formers for each of the microphone arrays configured by the microphone array selection means to obtain the null former output sounds;
2. The sound collection device according to claim 1, wherein the area sound extraction means selects the null form output sound having the smallest output from the plurality of null form output sounds for each frequency, and sets the selected null form output sound as the area sound. Computer,
a null-former processing means for calculating a plurality of null-formers that form blind spots outside a target area for each of a plurality of microphone arrays that can be configured by a combination of the microphones of the microphone group based on an acoustic signal supplied from the microphone group that is configured from three or more microphones, and acquiring a null-former output sound;
functioning as an area sound extraction means for acquiring an area sound obtained by extracting a sound of the target area based on the null former output sound of each of the null formers;
The area sound extraction means
a null former selection unit that selects the null former output sound having the smallest output from the plurality of null former output sounds for each frequency of the microphone array as the beamformer output sound of the microphone array;
a beamformer selection unit that selects the beamformer output sound having the smallest output from the plurality of beamformer output sounds as an area sound ,
The null former processing means acquires the null former output sounds from a first azimuth angle range from −π/2 to 0 and excluding the target area direction, and a second azimuth angle range from π/2 to 0 and excluding the target area direction, for each of the microphone arrays, by the same number.
A sound recording program characterized by: In the sound collection method performed by the sound collection device,
The sound collection device includes a null former processing means and an area sound extraction means,
The area sound extraction means includes a null former selection unit and a beam former selection unit,
the null former processing means calculates a plurality of null formers that form blind spots outside a target area for each of a plurality of microphone arrays that can be configured by a combination of the microphones of the microphone group based on an acoustic signal supplied from a microphone group consisting of three or more microphones, and acquires null former output sounds;
The area sound extraction means acquires an area sound obtained by extracting a sound of the target area based on the null former output sound of each of the null formers,
the null former selection unit selects, for each of the microphone arrays, the null former output sound having the smallest output from the plurality of null former output sounds for each frequency, as the beamformer output sound of the microphone array;
the beamformer selection unit selects the beamformer output sound having the smallest output from the plurality of beamformer output sounds as an area sound ;
The null former processing means acquires the null former output sounds from a first azimuth angle range from −π/2 to 0 and excluding the target area direction, and a second azimuth angle range from π/2 to 0 and excluding the target area direction, for each of the microphone arrays, by the same number.
A sound collection method comprising:

Images