JP2018056902A - Sound collecting device, program, and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To deal with fluctuation in a noise level while suppressing a musical noise or the like when performing an area sound collection.SOLUTION: The present invention relates to a sound collecting device. The sound collecting device of the present invention, includes: means of forming a directional pattern in a target area direction from a beamformer from an input signal; means of extracting a non-target area sound existed in the target area direction from an output of the beamformer; means of calculating a correction coefficient that corrects between a non-target area sound component included in the output of the beamformer and the extracted non-target area sound component; and means of correcting the extracted non-target area sound by the calculated correction coefficient, suppressing the non-target area sound component included in the output of the beamformer by using a phase inversion non-target area sound obtained by inversing a phase, and enhancing the target area sound.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a sound collection device, a program, and a method, and can be applied to a device that emphasizes sounds in a specific area and suppresses sounds in other areas, for example.

  There is a beam former (hereinafter referred to as BF) using a microphone array as a technique for separating and collecting only sound in a specific direction in an environment where a plurality of sound sources exist. BF is a technique for forming directivity using the time difference between signals reaching each microphone (see Non-Patent Document 1). BF is roughly divided into two types, an addition type and a subtraction type.

  In particular, the subtraction type BF has an advantage that directivity can be formed with a smaller number of microphones than the addition type BF.

  FIG. 6 is a block diagram showing a configuration related to a conventional subtractive BF.

  In the conventional subtraction type BF shown in FIG. 6, the number of microphones is two.

  The conventional subtractive BF first calculates the time difference between signals arriving at each microphone by sounds that are present in a target direction (hereinafter also referred to as “target sound”) by a delay device, and adds a delay to the target sound. Match the phase. In the conventional subtractor BF delay unit, the time difference is calculated by the following equation (1).

In the following formula (1), d is the distance between the microphones, c is the speed of sound, and τ _i is the delay amount. In the following equation (1), θ _L is an angle from a vertical direction to a target direction with respect to a straight line connecting the microphones.
τ _L = (dsin θ _L ) / c (1)

Here, when the blind spot exists in the direction of the first microphone with respect to the center of the first microphone and the second microphone, the delay unit in the conventional subtractive BF has the input signal x ₁ ( Delay processing is performed for t). Thereafter, the input signal x ₁ (t) subjected to the delay process is subjected to a subtraction process according to the equation (2).
m ₁ (t) = x ₂ (t) −x ₁ (t−τ _L ) (2)

The subtraction process in the conventional subtraction type BF can be similarly performed in the frequency domain, and in this case, the expression (2) is changed to the expression (3).

Here, when θ _L = ± π / 2, the formed directivity is cardioid unidirectional as shown in FIG. 7A, and when θ _L = 0, π, FIG. As shown in (B), the figure is bi-directional. Hereinafter, a filter that forms unidirectionality from an input signal is referred to as a unidirectional filter, and a filter that forms bidirectionality is referred to as a bidirectional filter.

Further, in a conventional sound collecting device, by using a spectral subtraction (hereinafter also referred to as “SS”), it is possible to form directivity that is strong against a blind spot of bi-directionality. The directivity by SS is formed at all frequencies or a designated frequency band according to the equation (4). In the equation (4), the input signal X1 of the _first microphone is used, but the same effect can be obtained with the input signal X2 of the _second microphone. Here, β is a coefficient for adjusting the strength of SS. If the value becomes negative during subtraction, flooring processing is performed in which 0 or the original value is replaced with a smaller value. In this method, sound existing in a direction other than the target direction (hereinafter also referred to as “non-target sound”) is extracted by a bi-directional filter, and the amplitude spectrum | M ₁ | of the extracted non-target sound is determined as the amplitude spectrum of the input signal | By subtracting from X ₁ |, the target sound | Y ₁ | can be emphasized. The amplitude spectra | X _1k | and | M _1k | for each frequency are calculated from the equations (5) and (6). Here, Re and Im represent a real part and an imaginary part, respectively, and k represents a frequency.

  When it is desired to collect only sound (hereinafter referred to as “target area sound”) existing in a specific area (within the sound collection target area), the directivity of the conventional subtractive BF is linearly formed. Therefore, by directing the directivity in the direction of the target area, even if the sound source exists outside the target area (hereinafter referred to as “non-target area sound”), all the sound sources existing in the direction of the target area are collected. Resulting in.

  Therefore, in Patent Document 1, a method (area sound collection) of collecting a target area sound by using a plurality of microphone arrays, directing directivity from different directions to the target area, and intersecting the directivity in the target area. is suggesting.

In the area sound collection method described in Patent Document 1, first, the power ratio of the target area sound included in the BF output of each microphone array is estimated and used as a correction coefficient. For example, when the area sound collection described in Patent Document 1 is performed using two microphone arrays, the correction coefficient for the target area sound power is calculated by the equation (7) or (8).

In equations (7) and (8), | Y _1k | is the amplitude spectrum of the BF output of the first microphone array, | Y _2k | is the amplitude spectrum of the BF output of the second microphone array, and m is the frequency bin total, alpha ₁ is a power correction factor for BF output. In the equations (7) and (8), mode represents the mode value and median represents the median value.

  In the area sound collection method described in Patent Document 1, each BF output is then corrected with a correction coefficient, and SS is performed to extract a non-target area sound that exists in the target area direction. In the area sound collection method described in Patent Document 1, the target area sound can be extracted by performing SS extraction on the extracted non-target area sound from the output of each BF.

In the area sound collection method described in Patent Document 1, when extracting the non-target area sound | N ₁ | existing in the target area direction as viewed from the first microphone array, SS is obtained by multiplying the BF output | Y ₂ | of the second microphone array by the power correction coefficient α ₁ from the BF output | Y ₁ | of the first microphone array.
| N ₁ | = | Y ₁ | −α ₁ | Y ₂ | (9)

Thereafter, in the area sound collection method described in Patent Document 1, the non-target area sound is SS extracted from the BF output according to the equation (10), and the target area sound is extracted. γ ₁ is a coefficient for changing the strength at the time of SS.
| Z ₁ | = | Y ₁ | −γ ₁ | N ₁ | (10)

JP 2014-72708 A

Asano Tadashi, "Acoustic Technology Series 16 Sound Array Signal Processing-Sound Source Localization / Tracking and Separation-", Acoustical Society of Japan, Corona, February 25, 2011

  However, in the area sound collection method of Patent Document 1, when the volume level of background noise or non-target area sound is large, if SS is performed during extraction of the target area sound, an irritating allophone called musical noise is caused by residual noise. May occur, and not only the non-target area sound but also the target area sound may be suppressed and the sound may be distorted. Therefore, in the area sound collection method of Patent Document 1, it is difficult to hear sound due to these effects, and smooth communication by sound may be hindered.

  Therefore, there is a demand for a sound collection device, program, and method that can suppress musical noise and the like when performing area sound collection and can cope with fluctuations in noise level.

  The sound collecting device according to the first aspect of the present invention includes (1) directivity forming means for forming directivity in the direction of a target area from an input signal by a beam former, and (2) directivity formed by the directivity forming means. A non-target area sound extracting means for extracting a non-target area sound existing in the direction of the target area; and (3) an amplitude spectrum of a non-target area sound component included in the output of the beamformer, and the non-target area sound extracting means. Correction coefficient calculating means for calculating a correction coefficient for correcting the amplitude spectrum of the extracted non-target area sound component; and (4) the correction coefficient calculating means for calculating the non-target area sound extracted by the non-target area sound extracting means. The corrected non-target area sound is acquired by correcting with the correction coefficient calculated in step 1, the phase of the acquired corrected non-target area sound is inverted to obtain the phase inverted non-target area sound, and the acquired phase Using Utatesumi non-target area sound, and having an object area sound enhancement means emphasizing purpose area sound suppressing non-target area sound component contained in the output of the beamformer.

  The sound collecting program of the second aspect of the present invention is formed by (1) directivity forming means for forming directivity in the direction of a target area by a beam former from an input signal, and (2) the directivity forming means. Non-target area sound extraction means for extracting non-target area sound existing in the direction of the target area due to directivity; (3) an amplitude spectrum of a non-target area sound component included in the output of the beamformer; and the non-target area sound Correction coefficient calculating means for calculating a correction coefficient for correcting the amplitude spectrum of the non-target area sound component extracted by the extracting means; and (4) correcting the non-target area sound extracted by the non-target area sound extracting means. A corrected non-target area sound is acquired by correcting with the correction coefficient calculated by the coefficient calculating means, and the phase of the acquired corrected non-target area sound is inverted to reverse the non-target area sound. Using the acquired phase-inverted non-target area sound, the non-target area sound component included in the beamformer output is suppressed to function as target area sound enhancement means for enhancing the target area sound. And

  According to a third aspect of the present invention, there is provided a sound collection method performed by the sound collection device, comprising: (1) directivity formation means, non-target area sound extraction means, correction coefficient calculation means, and target area sound enhancement means; Directivity forming means, non-target area sound extracting means, correction coefficient calculating means, and target area sound emphasizing means. (3) The directivity forming means has directivity in the direction of the target area by a beamformer from the input signal. (4) the non-target area sound extracting means extracts non-target area sound existing in the target area direction by the directivity formed by the directivity forming means, and (5) the correction coefficient calculating means Calculating a correction coefficient for correcting between the amplitude spectrum of the non-target area sound component included in the output of the beamformer and the amplitude spectrum of the non-target area sound component extracted by the non-target area sound extraction unit; ) The target area sound enhancement means acquires the corrected non-target area sound by correcting the non-target area sound extracted by the non-target area sound extraction means with the correction coefficient calculated by the correction coefficient calculation means. Reversing the phase of the corrected non-target area sound to obtain a phase-inverted non-target area sound, and using the acquired phase-inverted non-target area sound, the non-target area sound component included in the beamformer output is obtained. It is characterized by emphasizing the target area sound by suppressing it.

  ADVANTAGE OF THE INVENTION According to this invention, the musical noise etc. at the time of performing area sound collection can be suppressed, and it can respond to the fluctuation | variation of a noise level.

It is the block diagram shown about the functional structure of the sound collection device which concerns on 1st Embodiment. It is explanatory drawing shown about the structural example at the time of directivity by the beam former (BF) of two microphone arrays which concern on 1st Embodiment from the different direction to the target area. It is the graph shown about the example of the signal processed with the sound collection device which concerns on 1st Embodiment. It is explanatory drawing shown about the experimental result for confirming the performance of the sound collection device which concerns on 1st Embodiment. It is the block diagram shown about the functional structure of the sound collection device which concerns on 2nd Embodiment. It is the block diagram shown about the structure of the conventional sound collection device. It is explanatory drawing explaining an example of the directional characteristic formed with the conventional directivity filter.

(A) First Embodiment Hereinafter, a first embodiment of a sound collection device, a program, and a 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 the sound collection device 100 of this embodiment.

  The sound collection device 100 uses two microphone arrays MA (MA1, MA2) to perform target area sound collection processing for collecting a target area sound from a sound source in the target area.

  The microphone arrays MA1 and MA2 are arranged at any place in the air where the target area exists. For example, as shown in FIG. 2, the positions of the microphone arrays MA1 and MA2 with respect to the target area may be anywhere as long as the directivities overlap only in the target area, and may be arranged opposite to each other with the target area interposed therebetween. Each microphone array MA is composed of two or more microphones M, and an acoustic signal is collected by each microphone M. In this embodiment, description will be made assuming that two microphones M (M1, M2) that collect sound signals are arranged in each microphone array MA. That is, each microphone array MA constitutes a 2ch microphone array. The number of microphone arrays MA is not limited to two. When there are a plurality of target areas, it is necessary to arrange a number of microphone arrays MA that can cover all areas. In this embodiment, a directional microphone such as a shotgun microphone may be used instead of the microphone array.

  The sound collection device 100 includes a data input unit 1, a directivity forming unit 2, a delay correction unit 3, spatial coordinate data 4, a target area sound power correction coefficient calculation unit 5, a non-target area sound extraction unit 6, and a non-target area sound amplitude. A spectrum correction coefficient calculation unit 7 and a target area sound enhancement unit 8 are provided. Detailed processing of each functional block constituting the sound collection device 100 will be described later.

  The sound collection device 100 may be configured entirely by hardware (for example, a dedicated chip or the like), or may be partially or entirely configured as software (program). For example, the sound collection device 100 may be configured by installing a program (including the sound collection program of the embodiment) in a computer having a processor and a memory.

(A-2) Operation of First Embodiment Next, the operation (sound collection method according to the embodiment) of the sound collection device 100 of the first embodiment having the above-described configuration will be described.

  The data input unit 1 converts the acoustic signals collected by the microphone arrays MA1 and MA2 from analog signals to digital signals. And the data input part 1 performs the conversion process (For example, the process converted from a time domain to a frequency domain using a fast Fourier transform etc.) about the said digital signal.

  The directivity forming unit 2 extracts, for each microphone array MA, a non-target area sound that exists in a direction other than the target direction (for example, by a bi-directional filter), and extracts the amplitude spectrum of the extracted non-target area sound of the input signal. By subtracting from the amplitude spectrum, a sound having a directivity in the direction of the target area (BF output) is acquired. Specifically, the directivity forming unit 2 acquires, as a BF output, a sound in which directivity is formed in the direction of the target area by the BF according to the equation (4) for each microphone array. When the input signal is not a microphone array but a signal input from a directional microphone, the processing of the directivity forming unit 2 may be omitted and the input signal may be supplied to the subsequent stage as it is. Good.

  The delay correction unit 3 calculates and corrects a delay caused by a difference in distance between the target area and each microphone array. First, the position of the target area and the position of the microphone array are acquired from the spatial coordinate data 4, and the difference in the arrival time of the target area sound to each microphone array is calculated. Next, with reference to the microphone array arranged farthest from the target area, a delay is added so that the target area sound reaches all the microphone arrays simultaneously.

  The spatial coordinate data 4 holds the position information of all the target areas, the microphone arrays, and the microphones constituting the microphone arrays.

  The target area sound power correction coefficient calculation unit 5 calculates a correction coefficient for equalizing the power of the target area sound component included in each BF output (the BF output corrected by the delay correction unit 3) using Equation (7) or ( 8) Calculate according to the equation.

  The non-target area sound extraction unit 6 SS SS each BF output corrected by the correction coefficient calculated by the target area sound power correction coefficient calculation unit 5 according to the equation (9), and extracts the non-target area sound existing in the target area direction. To do.

The non-target area sound amplitude spectrum correction coefficient calculating unit 7 includes an amplitude spectrum of the non-target area sound component included in the BF output (each BF output corrected by the correction coefficient calculated by the target area sound power correction coefficient calculating unit 5), A correction coefficient (hereinafter referred to as “μ ₁ ”) for calculating the same amplitude spectrum of the non-target area sound component extracted by the non-target area sound extraction unit 6 is calculated. In other words, the correction coefficient μ ₁ is for correcting between the amplitude spectrum of the non-target area sound component included in the BF output and the amplitude spectrum of the non-target area sound component extracted by the non-target area sound extraction unit 6. It can be said that it is a coefficient. Specifically, the non-target area sound amplitude spectrum correction coefficient calculation unit 7 calculates the correction coefficient μ ₁ (non-target area sound amplitude based on the ratio of the amplitude spectrum of the BF output and the extracted amplitude spectrum of the non-target area sound. Spectral correction coefficient) is calculated.

The non-target area sound amplitude spectrum correction coefficient calculation unit 7 first obtains the ratio (ratio) of the amplitude spectrum for each frequency (each component) between the BF output and the extracted non-target area sound. Then, the non-target area sound amplitude spectrum correction coefficient calculation unit 7 calculates the mode value from the ratio of the amplitude spectrum for each frequency, and obtains the most frequent value as the correction coefficient μ ₁ .

Specifically, the non-target area sound amplitude spectrum correction coefficient calculation unit 7 can obtain μ ₁ using, for example, the equation (11).

In equation (11), | Y _1k | is the amplitude spectrum of the BF output of the microphone array MA1, | N _1k | is the amplitude spectrum of the extracted non-target area sound, k is the frequency, m is the total number of frequency bins, and mode is the maximum. Mode values (functions for obtaining the most frequent value) are shown. Further, in the expression (11), when | N _1k | = 0, | Y _1k | / | N _1k | is set to 0 to avoid 0%.

The non-target area sound extraction unit 6 extracts a non-target area sound component | N _1k | from the BF output | Y _1k |. That is, the BF output | Y _1k | includes the same non-target area sound components as the extracted | N _1k | with the same ratio and distribution. Therefore, when the ratio of the amplitude spectrum of the BF output | Y _1k | and the extracted non-target area sound | N _1k | (| Y _1k | / | N _1k |) is obtained, the ratio in the non-target area sound component is the same. Become. In contrast, since components other than the non-target area sound component (target area sound component, background noise component, etc.) are included only in the BF output | Y _1k |, the above ratio varies and becomes a large value. Based on the characteristics as described above, the non-target area sound amplitude spectrum correction coefficient calculation unit 7 calculates a ratio | Y _1k | / | N _1k | for each component (for each frequency), and sets a predetermined threshold (hereinafter, “Τ ₁ ), the components having a larger value ratio are eliminated, and then the mode value is obtained for the remaining component ratios. As a result, the non-target area sound amplitude spectrum correction coefficient calculation unit 7 determines that the calculated mode value is the non-target area sound in the BF output | Y _1k | and the amplitude spectrum of the extracted non-target area sound | N _1k | Are equal to the correction coefficient.

Note that the non-target area sound amplitude spectrum correction coefficient calculation unit 7 is not the mode (mode) of the ratio | Y _1k | / | N _1k | as in the equation (11), but as in the equation (12) As an approximation, the median may be used as the correction coefficient μ ₁ .

  The target area sound emphasizing unit 8 does not perform SS at the time of extracting the target area sound that may cause musical noise, and suppresses the non-target area sound from the BF output and emphasizes the target area sound using the phase information. .

  The phase information of the input signal can be obtained by dividing the real part and the imaginary part of the input signal subjected to the discrete Fourier transform by the amplitude spectrum of the input signal. The target area sound emphasizing unit 8 can have the same phase as the input signal by multiplying the phase information by the amplitude spectrum of the non-target area sound and the BF output. The phase information of the input signal originally includes the phase of the target sound in BF. Since the BF output includes the target area sound and the non-target area sound, if the phase information of the input signal is added to the BF output, the phase of the target area sound and the non-target area sound in the BF output can be restored. Can do. The target area sound emphasizing unit 8 can also restore the phase of the non-target area sound extracted in the same manner. From this, if the phase of the extracted non-target area sound is reversed and added to the BF output, the non-target area sound component in the BF output is canceled and suppressed with the extracted non-target area sound component. As a result, the target area sound emphasizing unit 8 can emphasize only the target area sound during BF output.

  Hereinafter, an example of specific processing of the target area sound enhancement unit 8 will be described.

The target area sound emphasizing unit 8 outputs the non-target area sound | N ₁ | (components for each frequency | N _1k | as a vector, the same _applies hereinafter) extracted by the non-target area sound extracting unit 6. It is corrected by the correction coefficient μ ₁ calculated by the non-target area sound amplitude spectrum correction coefficient calculation unit 7 and obtained as | N ′ ₁ |. Specifically, the target area sound enhancement unit 8 obtains | N ′ ₁ | by multiplying the non-target area sound | N ₁ | by the correction coefficient μ ₁ as shown in the equation (13).

Then, the target area sound emphasizing unit 8 gives phase information to | N ′ ₁ | according to the equation (14).

Then, the target area sound emphasizing unit 8 gives the phase information of the input signal to the BF output | Y ₁ | according to the equation (15).

The objective area sound enhancement unit 8, a band which is limited to the total bandwidth or a predetermined range in accordance with equation (16), by inverting _N'1 phase by summing the Y _1, non-in BF Output It suppresses the destination area sound components, to obtain the signal Z ₁ emphasizing the target area sound.

The objective area sound enhancement unit 8 outputs a signal Z ₁ emphasizing the target area sound (as such or, converts the time domain) to.

In the equations (13) to (16), μ ₁ is a correction coefficient calculated by the non-target area sound amplitude spectrum correction coefficient calculation unit 7. In the target area sound emphasizing unit 8, suppression of the non-target area sound may be performed in the frequency domain or may be performed after returning to the time domain by inverse Fourier transform.

  As described above, since the non-target area sound extracted in the area sound collection process is an amplitude spectrum and phase information is lost, the target area sound emphasizing unit 8 adds phase information of the input signal thereto. To do. Since the BF output is also an amplitude spectrum, the target area sound enhancement unit 8 similarly provides phase information of the input signal.

(A-3) Effects of First Embodiment According to this embodiment, the following effects can be achieved.

  (A-3-1) In the sound collection device 100 of the first embodiment, the non-target area sound is being suppressed without performing SS at the time of extracting the target area sound that may cause musical noise. The target area sound is emphasized by canceling out the extracted non-target area sound with an opposite phase to the non-target area sound component. As a result, the sound collection device 100 of the first embodiment has the effects of reducing the occurrence of musical noise and reducing the distortion of the target area sound by suppressing it.

  Usually, the amplitude spectrum of the non-target area sound extracted by the SS is smaller than the amplitude spectrum of the non-target area sound being output in the BF due to the influence of background noise or the like. In this state, even if the phase of the extracted non-target area sound is reversed and added to the BF output, it cannot be completely cancelled. On the other hand, when the process of canceling the non-target area sound with the opposite phase is performed as in the sound collecting apparatus 100 of the first embodiment, the process is not a non-linear process but a linear process as in SS. Therefore, generation of musical noise as in SS can be suppressed.

(A-3-2) In the sound collection device 100 of the first embodiment, the non-target area sound component included in the BF output and the non-target area sound extraction unit 6 extracted as the correction coefficient μ ₁ A sound component having the same amplitude spectrum is calculated and obtained (for example, calculated according to equation (11) or (12)).

For example, when the non-target area sound is suppressed with the correction coefficient μ ₁ as a constant, the amplitude spectrum of the non-target area sound component in the BF output and the non-target area sound component extracted from the BF output are the same size. It will be assumed. Also, for example, when the correction coefficient μ ₁ is set according to the background noise level, the change in the amplitude spectrum of the extracted non-target area sound is not always linear in an environment where the background noise level varies greatly. There is a possibility that it cannot follow. Also depending on the type of background noise, since the magnitude of the amplitude spectrum of the extracted non-target area sound changes, just the corresponding possible to proportion the correction coefficient mu ₁ to the level of background noise (variations in the appropriate correction coefficient mu ₁ Is difficult).

As described above, in the sound collection device 100 according to the first embodiment, the correction coefficient μ ₁ is obtained from the BF output and the amplitude spectrum ratio of the extracted non-target area sound, thereby changing the background noise level and the difference in type. Area pickup by robust phase suppression is possible.

  (A-3-3) Next, FIG. 3 and FIG. 4 are graphs of an experiment (hereinafter referred to as “main experiment”) that the applicant actually constructed and constructed the sound collection device 100 of the first embodiment. Will be described.

  FIG. 3 is a graph showing the waveform of the input signal processed by the target area sound enhancement unit 8 in this experiment. FIG. 3A is a graph showing the waveform of an input signal (input signal acquired by any microphone) input to the sound collection device 100. FIG. 3B is a graph showing the waveform of the target area sound included in the input signal shown in FIG. FIG. 3C is a graph showing a background noise waveform included in the input signal shown in FIG. The background noise shown in FIG. 3C is so-called office noise, and the power is increased by 6 db on the way.

  FIG. 4 is a graph showing the performance when the target area sound is emphasized by suppressing the non-target area sound component by using the sound collecting device 100 (target area sound emphasizing unit 8) of this embodiment.

FIG. 4A shows the target area by suppressing the non-target area sound component from the signal (BF output) based on the input signal as shown in FIG. 3A with the correction coefficient μ ₁ as a value proportional to the background noise. It is the graph shown about the waveform of the signal which emphasized the sound.

On the other hand, FIG. 4B uses the target area sound enhancement unit 8 of this embodiment to suppress the non-target area sound component from the signal (BF output) based on the input signal as shown in FIG. 6 is a graph showing a waveform of a signal in which a target area sound is emphasized. That is, the waveform shown in FIG. 4B is the correction coefficient μ calculated based on the above-described equation (11) or (12) from the signal (BF output) based on the input signal shown in FIG. _The result of canceling out the non-target area sound corrected in ₁ with the opposite phase is shown.

  From the experimental results shown in FIG. 4, it can be seen that the background noise is suppressed more in the signal in FIG. 4B than in the signal in FIG. In particular, in the signal of FIG. 4B, it can be seen that the background noise is suppressed more than the signal of FIG. 4A in the region where the background noise is larger (see FIG. 3C).

Therefore, from the experimental results shown in FIG. 4, it is more preferable to obtain the correction coefficient μ ₁ from the BF output and the amplitude spectrum ratio of the extracted non-target area sound than to set the correction coefficient μ ₁ to a value proportional to the background noise. It can be seen that the effect of suppressing noise is high.

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

(B-1) Configuration of Second Embodiment FIG. 5 is a block diagram showing a functional configuration of the sound collection device 100A of this embodiment. In FIG. 5, the same or corresponding parts as those in FIG.

  Hereinafter, differences from the second embodiment will be described for the sound collection device 100A of the second embodiment.

  The sound collection device 100A is different from the second embodiment in that the target area sound enhancement unit 8 is replaced with the target area sound enhancement unit 8A.

  The target area sound enhancement unit 8A enhances the target area sound by switching between suppression of the non-target area sound by the phase and suppression of the non-target area sound by the SS depending on the situation.

  The target area sound enhancement unit 8A includes a non-target area sound phase suppression unit 8-2 that performs non-target area sound suppression processing based on phase, and a non-target area sound spectrum subtraction unit 8 that performs non-target area sound suppression processing based on SS. −1 and switching to one of them to suppress non-target area sound (emphasis of target area sound).

(B-2) Operation of Second Embodiment Next, the operation of the sound collection device 100A of the second embodiment having the above configuration will be described.

  Below, only the difference with 2nd Embodiment is demonstrated about the sound collection apparatus 100A of 2nd Embodiment.

  The target area sound emphasizing unit 8A switches between the non-target area sound spectrum subtracting unit 8-1 and the non-target area sound phase suppressing unit 8-2, and suppresses the non-target area sound from the BF output.

The non-target area sound spectrum subtracting unit 8-1 performs processing for suppressing the non-target area sound by spectrum subtraction (SS) and emphasizing the target area sound. Specifically, the non-target area sound spectrum subtraction unit 8-1 suppresses the non-target area sound according to the equation (10). At this time, the non-target area sound spectrum subtraction unit 8-1 may use μ ₁ calculated by the non-target area sound amplitude spectrum correction coefficient calculation unit 7 instead of the coefficient γ ₁ for adjusting the intensity at the time of SS. Good.

  The non-target area sound phase suppression unit 8-2 performs processing for suppressing the non-target area sound and emphasizing the target area sound using the phase information, similarly to the target area sound enhancement unit 8 of the first embodiment. .

  Specifically, the non-target area sound phase suppression unit 8-2 suppresses the non-target area sound according to the equation (13).

  In the target area sound enhancement unit 8A, the non-target area sound suppression method (switching between the non-target area sound spectrum subtraction unit 8-1 and the non-target area sound phase suppression unit 8-2) is performed according to the value of the correction coefficient μ1. I do.

For example, when the correction coefficient μ ₁ is larger than the predetermined threshold value Τ ₂ (μ ₁ > T ₂ ), the target area sound enhancement unit 8A determines that the musical noise is also large due to the influence of the background noise, It is assumed that the non-target area sound is suppressed by the target area sound phase suppressing unit 8-2.

Also, the object area sound enhancement unit 8A, the correction coefficient mu ₁ is, for a given threshold value T ₂ or less (μ ₁ ≦ T _2), it is determined that the state in which musical noise is reduced due to the effect of background noise, non It is assumed that the non-target area sound is suppressed by the target area sound spectrum subtraction unit 8-1.

In the target area sound emphasizing unit 8A, the non-target area sound suppression method may be arbitrarily switched by a user operation instead of automatic determination based on the correction coefficient μ1 (threshold value T ₂ ). The target area sound emphasizing unit 8A may perform the processes of the non-target area sound spectrum subtracting unit 8-1 and the non-target area sound phase suppressing unit 8-2 at the same time, and may mix the respective outputs to obtain the final output.

(B-3) Effects of Second Embodiment According to the second embodiment, the following effects can be achieved.

  In the target area sound enhancement unit 8A of the sound collection device 100A of the second embodiment, an appropriate method (higher quality) is selected from the non-target area sound phase suppression unit 8-2 and the non-target area sound spectrum subtraction unit 8-1. A method capable of performing simple processing) is selected and applied to suppression of non-target area sounds (emphasis of target area sounds).

Specifically, in the second embodiment of the collection device 100A (destination area sound enhancement unit 8A), when the correction coefficient mu ₁ is large, by applying the non-target area sound phase suppressor 8-2, The generation of musical noise due to the leftover of noise and the suppression of target area sounds are reduced. In the second embodiment of the collection device 100A (destination area sound enhancement unit 8A), by applying the non-target area sound spectrum subtraction unit 8-1 when the correction coefficient mu ₁ is small, precisely non-target It is possible to suppress area sounds. Further, when the correction coefficient μ ₁ is small, it is possible to reduce the occurrence of musical noise and the suppression of the target area sound even if the target area sound suppression processing by the SS is performed. The non-target area sound can be suppressed with higher accuracy than the processing by the area sound phase suppressing unit 8-2.

(C) Other Embodiments The present invention is not limited to the above-described embodiments, and may include modified embodiments as exemplified below.

  (C-1) In the second embodiment, the target area sound enhancement unit 8A simultaneously performs the processing of the non-target area sound phase suppression unit 8-2 and the non-target area sound spectrum subtraction unit 8-1, and outputs each of them. May be mixed to obtain the final output.

DESCRIPTION OF SYMBOLS 100 ... Sound collecting device, 1 ... Data input part, 2 ... Directivity formation part, 3 ... Delay correction part, 4 ... Spatial coordinate data, 5 ... Target area sound power correction coefficient calculation part, 6 ... Non-target area sound extraction part 7, non-target area sound amplitude spectrum correction coefficient calculation unit, 8 ... target area sound enhancement unit.

Claims (8)

Directivity forming means for forming directivity in the direction of the target area by a beamformer from an input signal;
Non-target area sound extracting means for extracting non-target area sound existing in the target area direction due to directivity formed by the directivity forming means;
Correction coefficient calculation for calculating a correction coefficient for correcting between the amplitude spectrum of the non-target area sound component included in the output of the beam former and the amplitude spectrum of the non-target area sound component extracted by the non-target area sound extraction means Means,
The non-target area sound extracted by the non-target area sound extraction means is corrected with the correction coefficient calculated by the correction coefficient calculation means to obtain a corrected non-target area sound, and the phase of the acquired corrected non-target area sound is calculated. Inverted to acquire phase-inverted non-target area sound, and using the acquired phase-inverted non-target area sound, suppress the non-target area sound component included in the output of the beam former to enhance the target area sound A sound collection device comprising: target area sound enhancement means.   The target area sound enhancement means suppresses the non-target area sound component included in the output of the beamformer by adding the acquired phase-inverted non-target area sound to the output of the beamformer. The sound collecting device according to claim 1, wherein the sound collecting device is emphasized.   The correction coefficient calculating means corrects based on a ratio between the amplitude spectrum of the non-target area sound component included in the output of the beam former and the amplitude spectrum of the non-target area sound component extracted by the non-target area sound extraction means. The sound collecting device according to claim 1, wherein a coefficient is calculated.   The correction coefficient calculation means, for each component, a ratio between the amplitude spectrum of the non-target area sound component included in the output of the beamformer and the amplitude spectrum of the non-target area sound component extracted by the non-target area sound extraction means The sound collection device according to claim 3, wherein the correction coefficient is calculated based on the most popular value or the median value of the acquired ratios. The target area sound enhancement means includes:
First suppression means for suppressing the non-target area sound component included in the output of the beam former by using the acquired phase-inverted non-target area sound and enhancing the target area sound;
Second suppression means for suppressing the non-target area sound component of the beamformer output by spectral subtracting the signal based on the non-target area sound extracted by the non-target area sound extraction means from the output of the beamformer. And
The target area sound enhancement unit selects either the first suppression unit or the second suppression unit on the basis of the value of the correction coefficient acquired by the correction coefficient calculation unit, and outputs the beamformer output. The sound collection device according to claim 1, wherein a component of the non-target area sound is suppressed.   The target area sound enhancement means selects the first suppression means when the correction coefficient acquired by the correction coefficient calculation means is greater than a predetermined threshold, and the correction coefficient acquired by the correction coefficient calculation means 6. The sound collection device according to claim 5, wherein the second suppression unit is selected when the value is equal to or less than a threshold value. Computer
Directivity forming means for forming directivity in the direction of the target area by a beamformer from an input signal;
Non-target area sound extracting means for extracting non-target area sound existing in the target area direction due to directivity formed by the directivity forming means;
Correction coefficient calculation for calculating a correction coefficient for correcting between the amplitude spectrum of the non-target area sound component included in the output of the beam former and the amplitude spectrum of the non-target area sound component extracted by the non-target area sound extraction means Means,
The non-target area sound extracted by the non-target area sound extraction means is corrected with the correction coefficient calculated by the correction coefficient calculation means to obtain a corrected non-target area sound, and the phase of the acquired corrected non-target area sound is calculated. Inverted to acquire phase-inverted non-target area sound, and using the acquired phase-inverted non-target area sound, suppress the non-target area sound component included in the output of the beam former to enhance the target area sound A sound collection program which functions as a target area sound enhancement means. In the sound collection method performed by the sound collection device,
Directivity forming means, non-target area sound extraction means, correction coefficient calculation means, and target area sound enhancement means,
The directivity forming means forms directivity in the direction of the target area by a beamformer from an input signal,
The non-target area sound extracting means extracts non-target area sound existing in the target area direction due to the directivity formed by the directivity forming means,
The correction coefficient calculation means corrects the correction between the amplitude spectrum of the non-target area sound component included in the output of the beam former and the amplitude spectrum of the non-target area sound component extracted by the non-target area sound extraction means. Calculate the coefficient,
The target area sound enhancement unit corrects the non-target area sound extracted by the non-target area sound extraction unit with the correction coefficient calculated by the correction coefficient calculation unit to obtain a corrected non-target area sound, and the acquired correction The phase of the completed non-target area sound is inverted to acquire the phase-inverted non-target area sound, and the non-target area sound component included in the output of the beamformer is suppressed using the acquired phase-reversed non-target area sound. And picking up the target area sound.

Images