JP2016149612A - Microphone interval control device and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a microphone interval control device capable of rapidly controlling a microphone interval in response to the frequency characteristics of a noise component.SOLUTION: A microphone interval control device is based on fixedly arranged multiple microphones. Since a microphone interval is changed by the combination of two microphones among the multiple microphones, a microphone interval is selected by selecting two acquisition signals to be output to a next stage device. A feature amount expressing the frequency and largeness of change in the inclination direction of a signal waveform concerning the acquisition signal of the predetermined microphone is calculated in order to perform the selection. The feature amount reflects the level of the inclusion of a high range component. The two acquisition signals are selected in response to the calculated feature amount.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a microphone interval control device and a program, and can be applied to, for example, a case where noise components in input signals captured by a plurality of microphones are suppressed.

  As a technique for suppressing a noise component using a microphone array having a plurality of microphones, a technique described in Patent Document 1 can be cited.

  The technique described in Patent Document 1 assumes that signal processing is performed using a microphone array composed of two microphones whose positional relationship is fixed. However, conventionally, the positional relationship of microphones used for signal processing is not always fixed. For example, as shown in FIG. 9, there may be a case where a microphone array having three or more microphones is applied, and two microphones are selected from the microphone array based on some criteria to perform signal processing.

  By the way, the wider the interval between two microphones, the sharper the directivity of the low frequency component, while the high frequency component has an error due to the spatial sampling theorem. On the contrary, as the microphone interval is narrower, the directivity of the low frequency component becomes wider, but the directivity of the high frequency component becomes more accurate. As described above, the band in which the sharp directivity is formed changes according to the microphone interval. Therefore, when background noise suppression processing is performed using a microphone array, appropriate noise suppression processing can be realized by switching the microphone interval according to the frequency characteristics of the noise component. For example, if background noise is generated in which power is concentrated in the low frequency components, a microphone array with a wide microphone interval is applied to sharpen the directivity of the low frequency components. By appropriately reflecting the components, the noise suppression effect can be sufficiently exerted.

  As described above, in order to enhance the effect of the microphone array under various noises, the interval between the microphones constituting the microphone array is controlled according to the frequency characteristic of the noise component. Here, as a method for examining the frequency characteristics of the noise component, a method of extracting a frequency component of interest using a filter such as a low-pass filter or a high-pass filter has been conventionally employed.

JP 2013-1206026 A JP 2014-106337 A

Naofumi Aoki, “A Band Extension Technology for Narrow-Band Telephony Speech Base on Full Wave Rectification”, IEICE Trans. Commun. , Vol. E93-B (3), pp. 729-731, 2010

  However, in the conventional method of grasping the frequency characteristic of the noise component using the filter, (i) the filter process involves a processing delay, so that it cannot follow a sudden change in the noise characteristic. (Ii) The filter process is software-like. However, there is a problem that real-time performance is impaired due to an increase in the amount of computation, and costs are increased in order to secure necessary resources (for example, memory).

  Therefore, a microphone interval control device and a program that can quickly control the microphone interval according to the frequency characteristics of the noise component are desired.

  The microphone interval control apparatus according to the first aspect of the present invention is (1) the number of times and the magnitude of the change in the inclination direction of the signal waveform in the captured signal of the microphone based on at least one captured signal of the captured signals from the plurality of microphones And a feature amount calculating means for calculating a feature amount that reflects the degree of inclusion of the high frequency component, and (2) two captured signals to be output to the next-stage device according to the calculated feature amount. And a microphone interval optimizing means for adjusting a microphone interval between the output microphones.

  A microphone interval control program according to a second aspect of the present invention is a computer that controls (1) the number of times that the inclination direction of a signal waveform in a captured signal of a microphone changes based on at least one captured signal among captured signals from a plurality of microphones. And a feature amount calculation means for calculating a feature amount that reflects the degree of inclusion of the high-frequency component, and (2) two output to the next-stage device according to the calculated feature amount It is made to function as a microphone space | interval optimization means which adjusts the microphone space | interval between the microphones which output the capture signal.

  ADVANTAGE OF THE INVENTION According to this invention, the microphone space | interval control apparatus and program which can control a microphone space | interval rapidly according to the frequency characteristic of a noise component can be provided.

It is a block diagram which shows the structure of the microphone space | interval control apparatus of 1st Embodiment. It is a block diagram which shows the detailed structure of the microphone selection control part in the microphone space | interval control apparatus of 1st Embodiment. It is explanatory drawing which shows the structure of the modGI / selection control signal corresponding | compatible memory | storage part (table) in the microphone selection control part in the microphone space | interval control apparatus of 1st Embodiment. It is a block diagram which shows the structure of the microphone space | interval control apparatus of 2nd Embodiment. It is explanatory drawing which shows the structure of the modGI / selection / delay control signal corresponding | compatible memory | storage part (table) in the microphone selection / delay control part in the microphone space | interval control apparatus of 2nd Embodiment. It is a block diagram which shows the structure of the microphone space | interval control apparatus of 3rd Embodiment. It is explanatory drawing which shows the structure of the modGI / microphone position corresponding | compatible memory | storage part (table) in the microphone movement control part in the microphone space | interval control apparatus of 3rd Embodiment. It is a block diagram which shows the structure of the microphone space | interval control apparatus of 4th Embodiment. It is explanatory drawing of a microphone array.

(A) First Embodiment Hereinafter, a first embodiment of a microphone interval control device and a program according to the present invention will be described with reference to the drawings.

(A-1) Configuration of the First Embodiment FIG. 1 is a block diagram showing the configuration of the microphone interval control device 10 of the first embodiment, and the microphone capture signal output from the microphone interval control device 10 is The input microphone array processing unit 20 is also described.

  In the configuration of the microphone interval control device 10 shown in FIG. 1, the components other than the three microphones m1 and m2 may be constructed by connecting various hardware components, and the CPU Further, it may be constructed so as to realize the function by applying an execution configuration of a program such as a ROM or a RAM. Regardless of which construction method is applied, the functional detailed configuration of the microphone interval control device 10 is the configuration shown in FIG.

  In FIG. 1, the microphone interval control device 10 according to the first embodiment includes first to third microphones m1 to m3, a modGI calculation unit 11, a microphone selection control unit 12, and a microphone selection unit 13.

  Each of the first to third microphones m1 to m3 captures sound waves such as ambient sounds and sounds and converts them into electrical signals (analog signals). Each of the microphones m1 to m3 has non-directionality or directivity that is loose in the front direction described later. Although not shown in FIG. 1, an analog / digital converter is provided for converting captured signals (analog signals) from the microphones m1 to m3 into digital signals, and the captured signals s1 (n) converted into digital signals. To s3 (n). Here, “n” is a parameter representing time, but the description is omitted in the following description.

For the first embodiment, the first to third microphone m1~m3 are arranged in one straight line in this order, the microphone spacing _{D 12} between the first and second microphones m1 and m2, second and it is smaller than a microphone interval _{D 23} of the third microphone m2 and m3. Naturally, the microphone spacing _{D 13} between the first and third microphones m1 and m3 are and microphone distance _{D 12} of the first and second microphones m1 and m2, a microphone interval between the second and third microphones m2 and m3 wider than the D _23.

  Capture signals s1 to s3 from the first to third microphones m1 to m3 are given to the microphone selection unit 13. The microphone selection unit 13 has a configuration of three inputs and two outputs. In accordance with the selection control signal cntlmic output from the microphone selection control unit 12, two capture signals s1 to s3 input to the microphone selection unit 13 are captured. The signals SA and SB are selected and given to the microphone array processing unit 20. The microphone selection unit 13 includes, for example, two two-input one-output switches sw1 and sw2, and the first switch sw1 selects one of the captured signals s1 and s2, and the microphone as one captured signal SA after the selection. The second switch sw2 selects one of the captured signals s2 and s3, and supplies the selected signal to the microphone array processing unit 20 as the other captured signal SB after selection.

When the captured signals SA and SB given to the microphone array processing unit 20 are expressed by a combination of captured signals s1 to s3 before selection, (i) s1 and s2, (ii) s2 and s3, (iii) s1 and s3. is there. The intervals between the two microphones corresponding to the three types of combinations are D ₁₂ , D ₂₃ , and D ₁₃ , respectively. Therefore, it can be said that the microphone selection unit 13 selects the microphone interval.

In the case of the first embodiment, the captured signal s1 output from the first microphone m1 is given to the modGI calculation unit 11. The modGI calculation unit 11 calculates a modGI value modGI for the captured signal s1 and gives the obtained modGI value modGI to the microphone selection control unit 12. The mod GI calculation unit 11 calculates, for example, a mod GI value mod GI for the captured signal s 1 according to the equation (1).

  The modGI value will be briefly described (refer to Patent Document 2 for details). modGI means a modified gradient index (hereinafter referred to as GI). The GI before being corrected is described in Non-Patent Document 1.

  GI is an index for measuring the number of times and the magnitude of the change in the inclination direction of the signal waveform. The GI is obtained by dividing the sum of absolute difference values of successive samples when the tilt direction is changed by the square root of the power of the frame. Therefore, the GI is likely to increase as the number of changes in inclination within one frame increases, and also increases as the amount of change when the inclination changes increases. From such a property, it can be said that GI is directly connected to the amount of high frequency components included in the input waveform.

  However, since GI uses as a calculation element a variable ΔΨ (n) that takes only two values of 0 or 2 and has many jumps with large values in time series, the value increases or decreases irregularly. There is a characteristic ("value goes wild").

  In view of the fact that modGI has the property that the value of GI is rampant (has a jump with a large value), instead of GI, it has a high correlation with GI, and a new feature with stable changes that suppresses a large jump in value It is proposed as a quantity. modGI is the “power of the second-order difference of the calculation target signal” normalized by the “power of the calculation target signal” for an arbitrary signal (captured signal s1 in the present application) of the feature quantity calculation target (this is a constant). Doubled is also included).

  Since modGI has a high correlation with GI, it functions as a stable index that measures the number and magnitude of changes in the slope direction of the signal waveform, and reflects the amount of high frequency components included in the input waveform. Function.

  As described above, modGI is directly connected to the amount of high-frequency components included in the signal waveform (capture signal s1), so that the frequency characteristics of noise components can be estimated by observing modGI, and modGI Since the calculation is simple, it does not increase the cost. Therefore, modGI is used as an index (parameter) for estimating the frequency characteristic of the noise component.

  The expression (1) is the same calculation expression as the expression (13) in Patent Document 2, but the modGI value obtained by applying the expressions (5) and (10) to (12) described in Patent Document 2. You may make it calculate modGI.

  The microphone selection control unit 12 is based on the modGI value modGI of the acquisition signal s1 obtained by the modGI calculation unit 11, in other words, based on the frequency characteristics of the noise component in the acquisition signal s1, and from the two acquisition signals s1 to s3 to 2. A selection control signal cntlmic for selecting two captured signals is generated and provided to the microphone selection unit 13.

  FIG. 2 is a block diagram showing a detailed configuration of the microphone selection control unit 12. In FIG. 2, the microphone selection control unit 12 includes a modGI acquisition unit 21, a modGI collation / selection control signal generation unit 22, a modGI / selection control signal correspondence storage unit 23, and a selection control signal transmission unit 24.

  The modGI acquisition unit 21 takes in the modGI value modGI obtained by the modGI calculation unit 11. The modGI collation / selection control signal generation unit 22 obtains (generates) a selection control signal cntlmic by extracting the corresponding information in the modGI / selection control signal correspondence storage unit 23 using the acquired modGI value modGI as a key. . The modGI / selection control signal correspondence storage unit 23 stores a plurality of ranges of the modGI value modGI and selection control signals associated with the ranges. The selection control signal transmission unit 24 transmits the selection control signal cntlmic obtained by the modGI matching / selection control signal generation unit 22 to the microphone selection unit 13.

  FIG. 3 is an explanatory diagram illustrating a configuration example of the modGI / selection control signal correspondence storage unit 23, and FIG. 3 illustrates a table configuration. In the example of FIG. 3, the range of values that the modGI value modGI can take is divided into three using two boundary values α and β (β> α).

When the modGI value modGI acquired by the modGI acquisition unit 21 belongs to the smallest value range, that is, when modGI ≦ α, the captured signals s1 and s3 from the first and third microphones m1 and m3 are selected. The content of the selection control signal cntlmic is “10”. The spacing D ₁₃ between the first and third microphones m1 and m3 is the largest of the three intervals, the frequency characteristic of the noise component in the acquisition signal s1 is the preferred spacing when the low-frequency closer.

When the modGI value modGI acquired by the modGI acquisition unit 21 belongs to an intermediate range, that is, when α <modGI ≦ β, the captured signals s2 and s3 from the second and third microphones m2 and m3 are selected. The content of the selection control signal cntlmic is “01”. The spacing D ₂₃ of the second and third microphones m2 and m3 is in the middle of the three intervals, suitable intervals when the frequency characteristic of the noise component in the acquisition signal s1 is not a high-pass nearer the low range nearer It is.

Selection control for selecting captured signals s1 and s2 from the first and second microphones m1 and m2 when the modGI value modGI acquired by the modGI acquisition unit 21 belongs to the largest range, that is, when β <modGI. The content of the signal cntlmic is “11”. The spacing D ₁₂ between the first and second microphones m1 and m2 are the smallest of the three intervals are suitable distance when the frequency characteristic of the noise component of the high-frequency closer in the acquisition signal s1.

  The upper bit of the 2-bit selection control signal cntlmic defines an input terminal that is valid for the first switch sw1 of the microphone selection unit 13, and the upper bit “1” is the a input terminal of the first switch sw1. The upper bit “0” indicates that the b input terminal of the first switch sw1 is enabled. The lower bit of the 2-bit selection control signal cntlmic defines the input terminal that is effective for the second switch sw2 of the microphone selection unit 13, and the lower bit “1” is the a input terminal of the second switch sw2. And “0” in the lower bit indicates that the b input terminal of the second switch sw2 is enabled.

  The microphone array processing unit 20 performs processing based on a known technique on the two captured signals SA and SB that are input as input signals from a microphone array composed of two microphones. For example, when the microphone array processing unit 20 forms a directional signal as the output signal out, the microphone array processing unit 20 has an internal configuration as shown in FIG. As a result, an output signal out (n) is obtained.

(A-2) Operation of the First Embodiment Next, the operation of the microphone interval control apparatus 10 according to the first embodiment will be described. Below, it demonstrates in order of the whole operation | movement and the operation | movement in the microphone selection control part 12. FIG.

  The first to third captured signals s1 to s3 obtained by capturing the surrounding sound waves such as sound and sound by the first to third microphones m1 to m3 are given to the microphone selecting unit 13. The first acquisition signal S1 is given to the modGI calculation unit 11.

  The modGI value modGI for the first acquisition signal s 1 is calculated based on the equation (1) in the modGI calculation unit 11, and the obtained modGI value modGI is given to the microphone selection control unit 12. Then, the microphone selection control unit 12 selects two captured signals from the three captured signals s1 to s3 based on the calculated modGI value modGI, in other words, based on the frequency characteristics of the noise component in the captured signal s1. A selection control signal cntlmic is generated and supplied to the microphone selection unit 13.

  Of the three captured signals s1 to s3 from the first to third microphones m1 to m3, two captured signals SA and SB indicated by the selection control signal cntlmic are selected by the microphone selecting unit 13 and the microphone array processing unit. 20 is given.

  Then, in the microphone array processing unit 20, the two captured signals SA and SB are processed based on a known method as input signals from a microphone array composed of two microphones.

  Next, a detailed operation in the microphone selection control unit 12 will be described.

  The modGI value modGI obtained by the modGI calculation unit 11 is taken in by the modGI acquisition unit 21 and given to the modGI collation / selection control signal generation unit 22. The modGI collation / selection control signal generation unit 22 performs collation with the modGI / selection control signal correspondence storage unit 23 by using the captured modGI value modGI as a key, and according to the range to which the acquired modGI value modGI belongs. The selection control signal cntlmic (see FIG. 3) is extracted from the modGI / selection control signal correspondence storage unit 23. Then, the selection control signal cntlmic is transmitted to the microphone selection unit 13 by the selection control signal transmission unit 24.

(A-3) Effects of the First Embodiment As described above, according to the first embodiment, the frequency characteristics of the noise components in the captured signal s1 (in other words, the captured signals s1 to s3) by the modGI value modGI. Based on the result, two microphones suitable for noise component processing (for example, noise suppression) can be selected to form a microphone array, and a processing unit for processing a captured signal from the microphone array can be configured. Processing accuracy can be improved. Here, since the modGI value modGI is used as a parameter reflecting the frequency characteristic of the noise component, the processing delay is small and the calculation is easy.

  By applying the microphone interval control device 10 of the first embodiment to, for example, a communication terminal (such as a smartphone or a mobile phone terminal) or a video conference system, an effect such as improvement in call sound quality can be expected. By applying the microphone interval control device 10 of the embodiment to, for example, a car navigation system or a voice command input device of a home appliance, an effect such as improvement of voice recognition accuracy can be expected. For example, a smartphone is carried by a user in various environments, and background noise inside the car (background noise when the window is closed) of a car in which the user gets in is mainly low-frequency components compared to outside the vehicle. Is. In addition, the background noise characteristics in the vicinity of the voice command input device for the car navigation system include, for example, whether the traveling speed is fast or slow, whether an air conditioner is used, and further, an automobile window (especially a front row window). It varies greatly depending on various conditions such as whether the is open or closed. It is very effective to apply the microphone interval control device 10 of the first embodiment to such a device in which the characteristics of the background noise change variously.

(B) Second Embodiment Next, a second embodiment of the microphone interval control device and program according to the present invention will be described with reference to the drawings.

  FIG. 4 is a block diagram showing the configuration of the microphone interval control apparatus 10A of the second embodiment. The same and corresponding parts as those in FIG. 1 according to the first embodiment described above are assigned the same and corresponding reference numerals. It shows.

  In FIG. 4, the microphone interval control apparatus 10A of the second embodiment includes first to third microphones m1 to m3, a modGI calculation unit 11, a microphone selection / delay control unit 12A, a microphone selection unit 13, and an added delay selection unit. 14

In the case of the first embodiment described above, the first to third microphones m1 to m3 are arranged on one straight line in this order. In the case of the second embodiment, the three microphones m1 to m3 are not linear, and are fixedly arranged at the positions of the three vertices of a triangle (for example, a right triangle or an obtuse triangle) as shown in FIG. Yes. However, the relationship between the microphone intervals is the same as in the first embodiment. That is, the microphone spacing _{D 12} between the first and second microphones m1 and m2 are narrower than the microphone spacing _{D 23} of the second and third microphones m2 and m3, the microphone interval between the first and third microphones m1 and m3 D ₁₃ is wider than the microphone interval D ₁₂ between the first and second microphones m 1 and m 2 and the microphone interval D ₂₃ between the second and third microphones m 2 and m 3.

  Here, the direction of the sound wave (directivity) intended by the designer of a device (for example, a voice command input device) on which the microphone interval control device 10A is mounted is determined by the first and third microphones m1 and m3. Suppose that it is the orthogonal direction of the connecting line segment. In this case, since the three microphones m1 to m3 are fixedly arranged at the positions of the three vertices of the triangle, the selection is made when the first and third microphones m1 and m3 are selected as the two microphones. There is no time difference in the components from the direction intended by the designer in the latter two captured signals SA and SB, but when the first and second microphones m1 and m2 are selected as the two microphones, When the second and third microphones m2 and m3 are selected as the two microphones, the two captured signals SA and SB after the selection have components from directions intended by the designer, as shown in FIG. A time difference corresponding to the distance L is generated.

  Therefore, in the second embodiment, the provision delay selection unit 14 is provided. The provision delay selection unit 14 eliminates the time difference between the captured signals SA and SB from the two selected microphones according to the selection result of the two microphones from the three microphones m1 to m3.

  The detailed configuration of the grant delay selection unit 14 is not limited as long as the above-described function can be exhibited, but can be configured as shown in FIG. 4 as an example. In the following description, it is assumed that the delay amount (time) corresponding to the distance L is d. The addition delay selection unit 14 delays the capture signal SA by time d, a 2-input 1-output switch 14A2 that switches between passing the capture signal SA as it is or passing the delay signal from the delay unit 14A1. The delay unit 14B1 delays the capture signal SB by time d, and the 2-input 1-output switch 14B2 that switches between passing the capture signal SB as it is or passing the delay signal from the delay unit 14B1. The two switches 14A1 and 14B1 select the a input terminal when the control bit is “1”, and select the b input terminal when the control bit is “0”.

  When the first and third microphones m1 and m3 are selected as the two microphones, there is no time difference between the two captured signals SA and SB after selection. The two captured signals SA and SB from 13 may be supplied to the microphone array processing unit 20 without being delayed. Further, when the second and third microphones m2 and m3 are selected as the two microphones, there is a relationship in which the captured signal SB side advances by the time difference d between the two captured signals SA and SB after selection. The provision delay selection unit 14 may delay the capture signal SB by the time difference d and provide it to the microphone array processing unit 20. Further, when the first and second microphones m1 and m2 are selected as the two microphones, there is a relationship in which the capture signal SA side advances by the time difference d between the two capture signals SA and SB after selection. The provision delay selection unit 14 may delay the capture signal SA by the time difference d and provide it to the microphone array processing unit 20.

  The microphone selection / delay control unit 12A mod_GI / selection according to the first embodiment so that the delay control as described above can be realized with respect to the captured signal after selection of the microphone according to the selection of the microphone (captured signal). Instead of the control signal correspondence storage unit 23, a modGI / selection / delay control signal correspondence storage unit 23A is provided.

  FIG. 5 is an explanatory diagram illustrating a configuration example of the modGI / selection / delay control signal correspondence storage unit 23A, and corresponds to FIG. 3 of the first embodiment described above.

  The modGI / selection / delay control signal correspondence storage unit 23A also has two microphones (capture signals) from three microphones m1 to m3 (capture signals) according to the range to which the modGI value modGI acquired by the modGI acquisition unit 21 belongs. The selection control signal cntlmic for selecting can be obtained. In addition, the modGI / selection / delay control signal correspondence storage unit 23A selects according to the range to which the modGI value modGI acquired by the modGI acquisition unit 21 belongs, in other words, according to the combination of two selected microphones. A delay control signal cntldl that eliminates the time difference between the subsequent acquisition signals SA and SB can be obtained. The delay control signal cntldl controls the added delay selection unit 14 as described above.

  According to the second embodiment, the same effect as that of the first embodiment can be obtained. In addition, according to the second embodiment, even when there is a deviation in the timing of capturing sound waves from a predetermined direction among a plurality of microphones, the deviation is eliminated and the selected acquisition signal is processed by microphone array processing. Can be given to the department.

  As shown in FIG. 1 according to the first embodiment, even when a plurality of microphones are arranged on one straight line, when the predetermined direction is not the orthogonal direction of the straight line, as in the second embodiment. It is necessary to provide a means for eliminating a large shift in capture timing.

(C) Third Embodiment Next, a third embodiment of the microphone interval control apparatus and program according to the present invention will be described with reference to the drawings.

  FIG. 6 is a block diagram showing the configuration of the microphone interval control apparatus 10B of the third embodiment. The same and corresponding parts as those in FIG. 1 according to the first embodiment described above are assigned the same and corresponding reference numerals. It shows.

  In FIG. 6, the microphone interval control apparatus 10 </ b> B of the third embodiment includes two microphones mf, mm, modGI calculation unit 11, microphone movement control unit 12 </ b> B, and microphone movement mechanism 15.

  Of the two microphones mf and mm, one microphone mf is fixedly provided, and the other microphone mm is provided movably by the microphone movement control unit 12B. The moving direction of the microphone mm is a direction to approach or move away from the microphone mf that is fixedly installed. That is, in the case of the third embodiment, the number of microphones is two, but the interval between the two microphones mf and mm can be changed.

  The microphone moving mechanism 15 is a mechanism for moving the movable microphone mm, and its detailed configuration is not limited. An example is as follows. The microphone moving mechanism 15 includes a stepping motor as a drive source, and a pinion is attached to the rotation shaft of the stepping motor. A rack that engages with the pinion can be linearly moved by a guide, and a movable microphone mm is fixedly attached to a part of the side or back of the rack.

  The modGI calculation unit 11 calculates a modGI value modGI for the captured signal s1 output from the microphone mf, and gives the obtained modGI value modGI to the microphone movement control unit 12B.

  Based on the calculated modGI value modGI, in other words, based on the frequency characteristic of the noise component in the captured signal s1, the microphone movement control unit 12B optimizes the distance between the two microphones mf and mm. A movement amount is determined, and a movement control signal cntlmv is given to the microphone moving mechanism 15.

  As long as the above-described functions can be exhibited, the internal configuration of the microphone movement control unit 12B is not limited. An example is as follows. The microphone movement control unit 12B stores the current position 12B1 (information) of the microphone mm. Further, the microphone movement control unit 12B associates a range to which the modGI value modGI belongs with a position of the microphone mm suitable for the range, and a modGI / microphone position correspondence storage unit (table) configured as shown in FIG. 23B is built-in. The microphone movement control unit 12B extracts the position of the microphone mm suitable for the calculated modGI value modGI from the modGI / microphone position correspondence storage unit (table) 23B, and the difference between the extracted position and the current position 12B1 of the microphone mm. And a movement control signal cntlmv for moving the microphone mm to the position taken out from the current position 12B1 of the microphone mm is generated and given to the microphone moving mechanism 15.

  Also in the third embodiment, the modGI reflecting the frequency characteristic of the noise component in the captured signal is calculated, and the interval between the two microphones is controlled accordingly, so the same as in the first embodiment. There is an effect.

(D) Fourth Embodiment Next, a fourth embodiment of the microphone interval control device and program according to the present invention will be described with reference to the drawings.

  FIG. 8 is a block diagram showing the configuration of the microphone interval control apparatus 10C of the fourth embodiment. The same and corresponding parts as those in FIG. 1 according to the first embodiment described above are assigned the same and corresponding reference numerals. It shows.

  In FIG. 8, the microphone interval control device 10C according to the fourth embodiment includes a voice section detection unit in addition to the first to third microphones m1 to m3, the modGI calculation unit 11C, the microphone selection control unit 12, and the microphone selection unit 13. 16

  The voice section detection unit 16 detects whether or not the first captured signal s1 is a voice section by a known voice detection technique, and gives the detection result V to the modGI calculation unit 11C.

  The modGI calculation unit 11C of the fourth embodiment calculates the modGI value modGI for the captured signal s1 when the detection result V represents a non-speech interval, and immediately before the detection result V represents a speech interval. The present mode GI value mod GI is maintained, and the mod GI value mod GI updated only in the non-voice interval as described above is given to the microphone selection control unit 12.

  Therefore, in the fourth embodiment, the microphone interval is reviewed only in the non-voice section (in other words, the noise section).

  According to the fourth embodiment, since the mod GI value is updated only in the non-speech period and the selection of the microphone is controlled, a higher effect than in the first embodiment can be expected.

(E) Other Embodiments In each of the above embodiments, mod GI is applied. However, the number of times that the inclination direction of the signal waveform changes also in the GI before correction (see Equation (4) in Patent Document 2). Therefore, the GI may be applied instead of the modGI in each of the above embodiments.

  In each of the embodiments described above, the microphone interval is constantly reviewed. However, the frequency characteristics of the background noise are often steady, and the apparatus is arranged so that the microphone interval is reviewed temporarily or intermittently. It may be configured. For example, the review of the microphone interval may be activated only at the start of use of the apparatus, such as when the communication terminal is turned on, and may be stopped thereafter.

  In the first, second, and fourth embodiments in which the microphone is selected, the case where the number of microphones is three has been described, but it is needless to say that the number of microphones is not limited to three.

  In each of the above embodiments, a method for obtaining a modGI value to be given to various control units from one captured signal s1 has been described. However, a method for obtaining a modGI value to be given to various control units is not limited to this. For example, a modGI value may be calculated for each of the three types of captured signals s1, s2, and s3, and an average value, a median value, and the like of the obtained three modGI values may be used as a modGI value to be given to the control unit.

  In the third embodiment, one of the microphones is movable. However, both microphones may be movable.

  In the above description, the embodiment in which the selection of the microphone and the movement of the microphone are alternatives has been described. However, an embodiment in which the selection of the microphone and the movement of the microphone are combined may be configured. In short, it is only necessary that the interval between the microphones from which the two captured signals input to the microphone array are obtained can be controlled to the modGI value.

  In each of the above embodiments, the control signal is switched immediately when the modGI value belongs to the new range. However, the control signal may be switched only when the modGI value belongs to the new range for a predetermined number of times. .

  Some microphone array processing units have a delay unit therein (see FIG. 4 of Patent Document 1). The delay unit in the microphone array processing unit may be used as the delay unit in the added delay selection unit 14 in the second embodiment.

  In the above description, the audio signal is described as the target sound. However, the technical idea of the present invention can be applied to the case where the acoustic signal such as a mechanical sound is the target sound.

  In each of the above embodiments, an apparatus or a program for immediately processing a captured signal from the microphone array has been described. However, the present invention is also applied to a case where the captured signal from the microphone array is recorded on a recording medium and reproduced. be able to.

  m1 to m3, mf, mm,... microphone, 10, 10A, 10B, 10C... microphone interval control device, 11, 11C... modGI calculation unit, 12 .. microphone selection control unit, 12A .. microphone selection / delay control unit, 12B. Microphone movement control unit, 13 ... microphone selection unit, 14 ... grant delay selection unit, 15 ... microphone movement mechanism, 16 ... voice interval detection unit, 21 ... modGI acquisition unit, 22 ... modGI collation / selection control signal generation unit, 23 ... modGI / selection control signal correspondence storage unit, 23A... modGI / selection / delay control signal correspondence storage unit, 24... selection control signal transmission unit.

Claims (6)

Based on at least one of the signals captured by a plurality of microphones, it represents the number and magnitude of changes in the direction of the signal waveform slope in the captured signal of the microphone, and reflected the inclusion of high-frequency components. A feature amount calculating means for calculating a feature amount;
A microphone interval control device, comprising: a microphone interval optimizing unit that adjusts a microphone interval between microphones that output two captured signals to be output to the next-stage device according to the calculated feature amount.   The microphone interval optimizing means selects capture signals from two microphones according to the calculated feature amount from capture signals obtained from three or more microphones fixedly installed, and The microphone interval control apparatus according to claim 1, wherein the microphone interval is virtually changed.   2. The microphone according to claim 1, wherein the microphone interval optimization unit changes the microphone interval by moving at least one movable microphone to a position corresponding to the calculated feature amount. Interval control device. The feature amount calculating means includes:
A target sound section detector for detecting a target sound section in the captured signal;
The microphone interval according to any one of claims 1 to 3, further comprising: a feature amount calculation main body that recalculates the feature amount only when the detection result of the target sound interval detection unit is not the target sound interval. Control device.   5. The coefficient feature quantity calculating means calculates a value obtained by normalizing the power of the second-order difference of the capture signal with the power of the capture signal as a feature quantity. Target sound component detection device. Computer
Based on at least one of the signals captured by a plurality of microphones, it represents the number and magnitude of changes in the direction of the signal waveform slope in the captured signal of the microphone, and reflected the inclusion of high-frequency components. A feature amount calculating means for calculating a feature amount;
A microphone interval control program that functions as a microphone interval optimizing unit that adjusts a microphone interval between microphones that output two captured signals to be output to the next stage device in accordance with the calculated feature amount. .

Images