Sound source direction calibration device equipped with a sound wave shielding hood

Abstract

The present invention provides a sound wave shielding hood 8 that can suppress the reception of an external sound arriving from the outside of a direction detection range TA and thereby reduce erroneous detection due to a grating lobe resulting from the external sound and a sound source direction detecting device 1 including the sound wave shielding hood 8. The sound wave shielding hood 8 is configured to prevent an array sensor 3, which includes multiple microphones 9 for receiving sound waves, from receiving an external sound arriving from the outside of a preset direction detection range TA and is configured to shield at least one of the multiple microphones 9 from a sound wave arriving from a periphery OE of the direction detection range TA.

Description

Technical Field

[0001] The present invention relates to a sound wave shielding hood that shields sound waves arriving from outside the calibration range of the sound source direction, and a sound source direction calibration device including the sound wave shielding hood.

Background Art

[0002] There is known a sound source direction calibration device having an array sensor in which a plurality of microphones for receiving sound waves are arranged, and calibrating the direction in which a sound source exists based on the sound pressure signals of the sound waves received by the array sensor. In this type of device, sound waves arriving from outside a preset direction calibration range (hereinafter referred to as external sound) may also be received by the microphones. In that case, depending on the arrangement of the microphones in the array sensor, false calibration (sometimes referred to as a false signal) called a grating lobe occurs within the direction calibration range due to the regularity of the phase difference obtained for the external sound received by each microphone, which may cause incorrect calibration of the sound source direction. Therefore, it is desirable to shield external sound from the sound source direction calibration device.

[0003] Patent Document 1 describes an example of a collector configured to shield external sound. The collector includes a sound collecting hood (hereinafter simply referred to as a hood) having a sound reflecting inner wall in the shape of a parabolic surface, and a microphone disposed inside the hood with its sound receiving surface facing forward, that is, toward the opening side of the hood. When the opening of the hood is directed toward a predetermined sound source, the above-described collector receives the sound incident inside the hood by the microphone, and also shields the external sound arriving from the outside of the hood toward the microphone with the hood.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] The use of the above-described hood for each microphone of the array sensor is considered to be effective in shielding external noise from each microphone. However, when the frequency of the sound wave to be measured by the array sensor is high, the distance between the microphones constituting the array sensor is set narrow for the purpose of preventing the grating lobe from occurring in the vicinity of the sound source direction. Therefore, the installation space for attaching the hood becomes narrow, or there may be a competition for the installation space between adjacent hoods, making it difficult to attach the hood. Further, even if the hood can be attached, due to the narrow distance between the microphones, the opening of the hood cannot be made sufficiently large, and there is a possibility of being easily affected by the diffraction of sound waves. Specifically, when the opening of the hood is shorter than the wavelength of the external noise, there is a possibility that the external noise cannot be sufficiently shielded by the diffraction phenomenon of the sound wave.

[0006] On the other hand, in order to suppress the influence of external noise due to the diffraction phenomenon, if the distance between the microphones is set wide, the grating lobe due to the external noise is likely to occur within the azimuth calibration range, and conversely, the sound source direction is likely to be mis-calibrated.

[0007] Therefore, it is conceivable to cover the entire array sensor with a hood. By doing so, the hood can be attached to the array sensor without being affected by the arrangement of the microphones. Further, since the opening of the hood can be set large, the influence of external noise due to the diffraction phenomenon can be suppressed.

[0008] However, when setting the sizes such as the inner diameter of the hood and the height of the hood so that all the microphones in the array sensor can receive sound waves arriving from within the azimuth calibration range, it may not be possible to obtain a significant grating lobe suppression effect when performing azimuth calibration of the sound source. This is because, due to the relationship between the size of the hood, the arrival direction of external noise with respect to the array sensor, and the arrangement of each microphone, among the microphones constituting the array sensor, a large number of microphones can receive external noise.

[0009] The present invention has been made paying attention to the above technical problems, and an object thereof is to provide a sound wave shielding hood capable of suppressing the reception of external noise arriving from outside the azimuth calibration range and suppressing mis-calibration due to grating lobes caused by external noise, and a sound source azimuth calibration device provided with the sound wave shielding hood.

Means for Solving the Problems

[0010] The features of the present invention for solving such problems are as follows. [1] A sound wave shielding hood configured to suppress an array sensor having a plurality of microphones for receiving sound waves from receiving external noise arriving from outside a preset azimuth calibration range, and for at least one of the plurality of microphones, it is configured to shield sound waves arriving from the peripheral edge of the azimuth calibration range. [2] The sound wave shielding hood according to [1] above, which is configured to surround at least a part of the periphery of the array sensor. [3] The sound wave shielding hood according to [1] or [2] above, which has a cylindrical shape, and the height protruding from the array sensor is set to be equal to or higher than the height for shielding sound waves arriving from the peripheral edge of the azimuth calibration range with respect to at least one of the plurality of microphones. [4] The height protruding from the array sensor is set lower than the height at which the number of microphones that are not shielded from sound waves arriving from the peripheral portion of the azimuth calibration range among the plurality of microphones becomes two, and it is the shielding hood according to any one of [1] to [3] above. [5] An array sensor having a plurality of microphones for receiving sound waves, based on the sound pressure information of the sound waves received by the array sensor, calculates the sound pressure of each azimuth within the azimuth calibration range respectively, and among each azimuth, an operation means for calibrating the azimuth with the maximum sound pressure as the arrival azimuth of the sound wave from the sound source, and a sound wave shielding hood according to any one of [1] to [4] above. It is a sound source azimuth calibration device having. [6] The average interval d between the plurality of adjacent microphones satisfies the following formula, and it is the sound source azimuth calibration device according to [5] above. d≧90°×λ / (90°+βrange / 2)(mm) In the above, λ is the center wavelength (mm) of the sound wave used for calculating the arrival azimuth of the sound wave, and βrange is the maximum angle (°) of the azimuth calibration range. [7] The operation means weights the sound pressure information of the sound waves received by each of the plurality of microphones according to the height of the sound wave shielding hood, and is configured to calibrate the azimuth of the sound source based on the weighted sound pressure information. It is the sound source azimuth calibration device according to [5] or [6] above. [8] The operation means performs delay-and-sum beamforming calculation on the plurality of sound pressure information obtained by each of the plurality of microphones to calibrate the arrival azimuth of the sound wave, and is the sound source azimuth calibration according to any one of [5] to [7] above. It is a device.

Effect of the Invention

[0011] According to the present invention, the sound wave shielding hood is installed so as to surround the array sensor, and for at least one of the plurality of microphones, sound waves and external sounds arriving from the peripheral portion of the azimuth calibration range are shielded. Therefore, as a whole of the device, reception of external sounds is sufficiently suppressed, and generation of grating lobes caused by external sounds can be suppressed. Thereby, mis-calibration of the azimuth of the sound source due to the grating lobes caused by external sounds can be suppressed. As a result, the azimuth calibration accuracy of the sound source by the sound source azimuth calibration device can be improved. Further, compared with the case where a sound wave shielding hood is provided for each microphone, the number of parts and the member cost can be reduced for the whole device, so an increase in the product cost can be suppressed.

Brief Description of the Drawings

[0012]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

[0013] Hereinafter, embodiments of the present invention will be described. FIG. 1 is a diagram showing an example of a sound source azimuth calibration device according to an embodiment of the present invention. FIG. 1(a) is a front view of the sound source azimuth calibration device, FIG. 1(b) is a cross-sectional view taken along the line A-A shown in FIG. 1(a), and FIG. 1(c) is a rear view of the sound source azimuth calibration device. FIG. 2 is a perspective view of the sound source azimuth calibration device according to the embodiment of the present invention. The sound source azimuth calibration device 1 according to the present embodiment shown here mainly includes a camera 2, an array sensor 3, an arithmetic unit 4, a display unit 5, an input unit 6, a housing 7, and a sound wave shielding hood (hereinafter simply referred to as a hood) 8. In the example shown in FIGS. 1 and 2, the camera 2, the array sensor 3, and the hood 8 are provided on the front side of the housing 7 of the sound source azimuth calibration device 1. Further, the arithmetic unit 4 is provided inside the housing 7. The display unit 5 and the input unit 6 are provided on the rear side of the housing 7.

[0014] The camera 2 captures an object to be measured and outputs the obtained captured image to the arithmetic means 4. The captured image is used for the purpose of superimposing it on the two-dimensional distribution of sound pressure (referred to as a sound pressure map) obtained by delay-and-sum beamforming calculation. Here, the delay-and-sum beamforming calculation refers to a calculation method for obtaining the arrival direction angle of a sound wave with respect to the front direction of the microphone by using the phase difference of the sound waves measured by a plurality of microphones in the array sensor 3. The magnification, field of view width, etc. of the camera 2 may be appropriately adjusted according to the search range of the sound source. The camera 2 may be, for example, a digital camera including an image sensor such as a CCD sensor or a CMOS sensor and a lens.

[0015] The above-described camera 2 is provided at the center of the array sensor 3. By providing the camera 2 at the center of the array sensor 3, when superimposing the captured image on the sound pressure map obtained by delay-and-sum beamforming calculation, an image can be created in which the captured image and the sound pressure map are superimposed without deviation regardless of the distance to the object to be measured.

[0016] The array sensor 3 has a plurality of microphones 9. The microphone 9 is configured to receive sound waves emitted from a sound source (not shown) and output sound pressure information of the received sound waves. In an embodiment of the present invention, each microphone 9 is provided on a flat base plate. FIG. 3 is a diagram showing an example of the arrangement pattern of the microphones 9. In the example shown in FIG. 3, on the same plane, each microphone 9 is arranged at each vertex of two concentric polygons (not shown) having the same center so as to be non-rotationally symmetric as a whole. The microphone 9 is configured to receive sound waves in the range from audible sound to a frequency of 100 kHz. The sound pressure information output from the microphone 9 is input to the arithmetic means 4, and is subjected to band-pass filter processing in an arbitrary frequency band by the arithmetic means 4, and delay-and-sum beamforming calculation is performed. In the embodiment of the present invention, although the case where the microphone 9 detects a predetermined broadband sound wave is illustrated, a single wavelength or a narrow-band sound wave may be detected.

[0017] In the example shown in FIG. 3, the array sensor 3 has 13 microphones 9. These microphones 9 are provided, for example, at the positions of the vertices of an outer regular heptagon with an outer circumscribed circle diameter of 41 mm and at the positions of the vertices of an inner regular hexagon with an outer circumscribed circle diameter of 15.3 mm when a 40 kHz sound wave (at a temperature of 25° C. and 1 atmosphere, with a wavelength of approximately 8.7 mm) propagating in the air is the measurement target. The outer circumscribed circle of the outer regular heptagon and the outer circumscribed circle of the inner regular hexagon are provided on concentric circles centered on the camera 2. Note that the frequency of the sound wave to be measured and the size of the outer circumscribed circle of the regular n-sided polygon are not limited to this. For example, the measurement frequency by the microphone 9 can be selected from within the range from audible sound (e.g., 20 Hz) to 100 kHz, and the size of the outer circumscribed circle of the regular n-sided polygon can be appropriately set according to that frequency (wavelength).

[0018] In the sound source direction calibration device 1 according to the embodiment of the present invention, as described above, 13 microphones 9 are provided at the respective vertices of an outer regular heptagon, which is a concentric polygon with the same center, and at the respective vertices of an inner regular hexagon. The microphones 9 arranged in this way are non-rotationally symmetric as a whole. Here, non-rotationally symmetric means an arrangement in which all the microphones 9 do not simultaneously return to the same arrangement as before rotation while rotating the array sensor 3 by 360° with respect to the center of the outer circumscribed circle of the concentric polygon. In this way, by arranging the plurality of microphones 9 to be non-rotationally symmetric as a whole, it is possible to improve the directivity while suppressing an increase in the number of installed microphones 9, and the array sensor 3 can be made to suppress the generation of grating lobes.

[0019] The display means 5 may be, for example, a liquid crystal display (LCD), and is configured to display the superimposed image created by the arithmetic means 4. The input means 6 may be, for example, push switches, and a plurality of them are provided in the vicinity of the display means 5. When the user presses the input means 6, a predetermined input signal is input to the arithmetic means 4. Note that instead of the push switches, or together with the push switches, a touch panel type input means 6 may be used.

[0020] FIG. 4 is a functional block diagram of the sound source direction calibration device 1. Using FIG. 4, the display process of the superimposed image in the sound source direction calibration device 1 will be described. The arithmetic means 4 includes a processing unit 10 and a storage unit 11. The processing unit 10 is, for example, a CPU or the like, and executes a predetermined calculation using the programs, data, acquired information, etc. stored in the storage unit 11, and outputs a control command signal for controlling the operation of the sound source direction calibration device 1. The storage unit 11 is, for example, an information recording medium such as a rewritable flash memory, a hard disk built-in or connected by a data communication terminal, a memory card, and its reading and writing device. Programs for realizing various functions of the sound source direction calibration device 1 and information used during the execution of the programs are stored in the storage unit 11 in advance.

[0021] The processing unit 10 uses the sound pressure information acquired from the array sensor 3 to perform delay-and-sum beamforming calculation for each predetermined azimuth division section in the azimuth calibration range, for example, based on the time difference (phase difference) of the sound pressure information detected by each microphone 9. By doing so, a sound pressure map in which the sound pressures of each azimuth division section are summarized in a two-dimensional distribution is created. The processing unit 10 creates a superimposed image in which the sound pressure map corresponding to the imaging area of the camera 2 is superimposed on the captured image. Specifically, the processing unit 10 converts the sound pressure map into a transparent image or a semi-transparent image, and creates a superimposed image by superimposing this on the captured image. In addition, the processing unit 10 outputs the superimposed image to the display means 5 for display. In the superimposed image, the subject and the sound pressure distribution are superimposed and displayed. Therefore, by the user checking the superimposed image, the correspondence between the position of the subject in the captured image and the sound pressure distribution of the sound wave can be discriminated at a glance. Therefore, in addition to calibrating the azimuth at which the sound pressure is maximum as the arrival azimuth of the sound wave, it is possible to determine which part of the subject the sound source is located at.

[0022] FIG. 5 is a diagram showing an example of the sound source azimuth calibration range TA by the sound source azimuth calibration device 1. In the example shown in FIG. 5, the azimuth calibration range TA forms a quadrilateral, which is predetermined in terms of design. Therefore, in the embodiment of the present invention, as a result, within the quadrangular pyramid-shaped region shown in FIG. 5 in front of the sound source azimuth calibration device 1, a search for a sound source (not shown) is performed. Also, the four sides of the azimuth calibration range TA shown in FIG. 5, that is, the outer peripheral surface of the bottom surface of the quadrangular pyramid and the azimuth corresponding to the vicinity inside thereof, correspond to the peripheral edge OE of the azimuth calibration range TA in the embodiment of the present invention. The maximum angle βrange of the azimuth calibration range TA in the embodiment of the present invention means the angle between the hypotenuses that are diagonally located on the bottom surface of the quadrangular pyramid among the four hypotenuses of the quadrangular pyramid shown in FIG. 5.

[0023] FIG. 6 is a diagram for explaining the arrival direction of a sound wave. As shown in FIG. 6, the arrival direction of the sound wave is the direction where the sound source exists, and is expressed as the arrival direction angle (θ, ψ) by the azimuth angle θ with respect to the azimuth (Z azimuth) orthogonal to the arrangement plane (XY plane) of the microphone 9 and the azimuth angle φ with respect to an arbitrary azimuth (for example, X azimuth) serving as a reference on the arrangement plane (XY plane). The processing unit 10 determines the arrival direction angle (θ, ψ) by delay-and-sum beamforming calculation using the sound pressure information detected by the plurality of microphones 9.

[0024] FIG. 7 is a schematic diagram showing how a sound wave is detected by a plurality of microphones 9, and FIG. 8 is a diagram showing an example of the sound pressure information detected by each microphone 9. For simplicity of explanation, FIGS. 7 and 8 illustrate the case where three microphones 9a, 9b, and 9c are arranged in a row at equal intervals. When the sound wave arrives from the above arrival direction angle (θ, ψ), when the microphone 9a is used as a reference microphone (hereinafter referred to as the first microphone 9a), the sound wave arrives at the second microphone 9b adjacent to the first microphone 9a with a time difference τ with respect to the first microphone 9a. The time difference τ (s) at which the sound wave arrives at the first microphone 9a and the second microphone 9b can be obtained by the following formula (1). τ=(L×sinθ) / v ···(1) In the above formula (1), v is the speed of sound (mm / s), θ is the azimuth angle of the sound wave, and L is the distance between the first microphone 9a and the second microphone 9b (the distance d between the microphones 9 to be described later) (mm).

[0025] In FIGS. 7 and 8, although it shows that a time difference τ (phase difference) occurs due to the azimuth angle θ, a time difference also occurs due to the azimuth angle ψ. The concept for the azimuth angle ψ is the same as in formula (1), and the time difference τ changes according to the arrival azimuth angles (θ, ψ).

[0026] Here, consider the case where sound waves arrive from a sound source located at a sufficient distance from each of the microphones 9a, 9b, 9c. The distance 2d between the first microphone 9a, which is the reference, and the third microphone 9c is set to twice the distance d between the first microphone 9a and the second microphone 9b. Therefore, the sound wave reaches the third microphone 9c with a time delay of τ compared to the second microphone 9b. That is, the time difference 2τ between the arrival time of the sound wave at the first microphone 9a and the arrival time of the sound wave at the third microphone 9c is twice the time difference τ between the arrival time of the sound wave at the first microphone 9a and the arrival time of the sound wave at the second microphone 9b. The sound wave reaches the third microphone 9c with a time delay twice that of the second microphone 9b. Therefore, as shown in FIG. 8, the sound pressure information detected by each of the microphones 9a, 9b, 9c is shifted by the time difference τ corresponding to the distance L from the reference first microphone 9a. Using these, the processing unit 10 calibrates the arrival azimuth angles (θ, ψ) of the sound waves.

[0027] Specifically, the processing unit 10 shifts the time of the sound pressure information received by each of the plurality of microphones 9 by the time difference τ, and calculates an addition value obtained by adding the plurality of time-shifted sound pressure information. A plurality of different time differences τ corresponding to the arrival azimuth angles (θ, ψ) are preset in the processing unit 10, and these plurality of different time differences τ are set for each of the plurality of microphones 9 according to the distance L. The processing unit 10 uses the plurality of different preset time differences τ to obtain a plurality of addition values by changing the time to be shifted as described above.

[0028] The added value obtained by adding with the time difference τ corresponding to the arrival azimuth angles (θ, ψ) becomes maximum because the phases of the waveforms of each sound pressure information are aligned. On the other hand, among the plurality of added values, the added value obtained by adding with the time difference τ that does not correspond to the arrival azimuth angles (θ, ψ) does not increase significantly because the phases are not aligned and they cancel each other out. Therefore, the processing unit 10 sequentially changes the assumed arrival azimuth angles (θ‘, ψ’) of the sound waves from the assumed sound sources assumed within the azimuth determination range TA of the sound source to obtain a plurality of added values. Further, using the plurality of added values thus obtained, a sound pressure map which is a two-dimensional distribution of the sound pressure within the azimuth determination range TA is created. Then, among the plurality of added values, the assumed arrival azimuth angle that becomes the maximum added value is determined as the azimuth of the sound source.

[0029] In the above delay-and-sum beamforming calculation, due to the relationship between the arrangement of each microphone 9 constituting the array sensor 3 and the sound wave wavelength, grating lobes are generated, which may cause mis-determination.

[0030] In addition, in the embodiment of the present invention, although the delay-and-sum beamforming calculation in which the processing unit 10 determines the arrival azimuth angles (θ, ψ) based on the time difference τ is illustrated, it is not limited thereto and various known methods can be applied. For example, the sound pressure information observed by each of the microphones 9a, 9b, 9c at the same time is regarded as a two-dimensional spatial sound pressure field. Then, by performing a two-dimensional Fourier transform spatially on the sound pressure information, a sound source azimuth determination result almost equivalent to the delay-and-sum beamforming calculation can be obtained by utilizing the correlation between the spatial wave number obtained and the arrival azimuth angles (θ, ψ).

[0031] As described above, the processing unit 10 is configured to display the superimposed image on the display means 5. As the sound pressure map used for the superimposed image displayed on the display means 5, a color image in which colors are continuously differentiated according to the sound pressure may be used. The processing unit 10 is configured to store the camera image and the superimposed image in the storage unit 11 when an instruction to save the image displayed on the display means 5 from the input means 6 is input.

[0032] In the sound source direction calibration device 1 according to this embodiment, from the viewpoint of improving the sound source direction calibration accuracy, it is preferable to use an array sensor 3 with high sound source direction calibration performance. The sound source direction calibration performance of the array sensor 3 is greatly affected by the arrangement of a plurality of microphones 9.

[0033] One of the sound source direction calibration performance indicators is directivity. Directivity is an indicator that shows whether the sound source sound pressure observed in the sound pressure map (hereinafter referred to as the main lobe) can be clearly displayed. The narrower the full width at half maximum (FWHM) of the main lobe, the higher the directivity, which is advantageous for calibrating the sound source direction. The full width at half maximum (FWHM) of the main lobe is represented by the ratio λ / D (rad) of the wavelength λ of the sound wave detected by the array sensor 3 to the aperture width D of the array sensor 3. To increase the directivity, it is effective to increase the aperture width of the array sensor 3 with respect to the wavelength. The aperture width of the array sensor 3 is the diameter of a virtual circle passing through a plurality of microphones 9 arranged on the outermost side with respect to the center of the array sensor 3. In the embodiment of the present invention, it is the diameter of the circumscribed circle that touches the outer co-circular polygon among the two co-circular polygons described above.

[0034] Another sound source direction calibration performance indicator is the sound source intensity ratio. The sound source intensity ratio is the ratio of the intensity of the grating lobe to the intensity of the main lobe, and is expressed by dividing the intensity of the main lobe by the intensity of the grating lobe. A high sound source intensity ratio means that the intensity of the grating lobe is sufficiently small compared to the intensity of the main lobe. If the intensity of the grating lobe can be made smaller than the intensity of the main lobe, misjudgment of the sound source direction due to the grating lobe can be suppressed, so a high sound source intensity ratio is advantageous for calibrating the sound source direction. To increase the sound source intensity ratio, it is effective to increase the number of microphones 9 and narrow the arrangement interval of the microphones 9. However, from the viewpoints of reducing product costs and reducing the calculation load on the calculation means 4, it is preferable that the number of microphones 9 is as small as possible. Therefore, as in the embodiment of the present invention, the generation of the grating lobe may be suppressed by devising the arrangement of the microphones 9.

[0035] Here, the hood 8 in the embodiment of the present invention will be described. As shown in FIGS. 1 and 2, the hood 8 is configured to be located on the arrangement plane of the array sensor 3 and outside the array sensor 3 so as to surround the array sensor 3. In the example shown in FIGS. 1 and 2, the hood 8 is formed in a square tube shape. Note that the hood 8 is not limited to a square tube. FIG. 9 is a schematic diagram showing another example of the shape of the hood 8. The shape of the hood 8 may be formed in a cylindrical shape as shown in FIG. 9(a), or may be formed in a truncated conical shell shape as shown in FIG. 9(b). Alternatively, it may be formed in an elliptical cylindrical shell shape as shown in FIG. 9(c), or may be formed in a truncated elliptical shell shape as shown in FIG. 9(d). Or, it may be formed in a truncated square pyramid shell shape as shown in FIG. 9(e).

[0036] The shape of the hood 8 does not necessarily have to be a truncated cone shape in which the observation direction side expands more than the microphone 9 side, that is, a tapered shape. That is, it may be a truncated cone shape in which the microphone 9 side expands more than the observation direction side. When the shape of the hood 8 is the above-described truncated cone shape in which the observation direction side expands more than the microphone 9 side, it is preferable that the expansion angle (taper angle) of the hood 8 does not expand to an angle (βrange / 2) that is half of the maximum angle of the azimuth calibration range TA. This is because when the taper angle of the hood 8 becomes the above-described angle or more, the sound waves arriving from the peripheral edge OE of the azimuth calibration range TA cannot be blocked by the hood 8, and as a result, sufficient external sound shielding performance may not be obtained. In the example shown in FIGS. 1 and 2, the hood 8 is formed in a square tube shape, but instead, it may be formed in a polygonal tube shape other than a square tube. Also, instead of the truncated square pyramid shell shape shown in FIG. 9(e), it may be formed in a polygonal truncated pyramid shell shape other than the truncated square pyramid shell shape.

[0037] Also, in the circumferential direction of the virtual circle described above, slits extending in the axial direction of the hood 8 may be formed at regular intervals in the hood 8. Alternatively, a plurality of plates (not shown) extending in the axial direction at regular intervals in the circumferential direction of the virtual circle described above may be arranged as shielding portions to form the hood 8. In short, the hood 8 should be located outside the array sensor 3 and configured to cover at least a part of the periphery of the array sensor 3 from the outside in the radial direction of the virtual circle described above.

[0038] The hood 8 in the embodiment of the present invention may be composed of a sound insulation material having sound insulation performance, a sound absorption material having sound absorption performance, or a combination thereof. The sound insulation material is composed of, for example, a synthetic resin material, a metal material, etc., and the sound absorption material is composed of soft urethane, polystyrene foam, melamine foam, rubber sponge type, fibrous glass wool, white wool, etc. When the hood 8 is configured by combining the sound insulation material and the sound absorption material, it is preferable to configure the inner part of the hood 8, that is, the part on the microphone 9 side in the plate thickness direction, with the sound absorption material. This is to prevent or suppress the multiple reflection of sound waves in the hood 8.

[0039] The size such as the height of the hood 8 and the inner diameter of the hood 8 described above is determined based on the maximum angle βrange of the orientation determination range TA of the sound source orientation determination device 1 and the arrangement pattern of each microphone 9 of the array sensor 3. Specifically, for example, the case where each microphone 9 is arranged as shown in FIG. 10 will be described as an example. Note that in the example shown in FIG. 10, for the sake of simplifying the explanation, the description of the sound source existing outside the orientation determination range TA is omitted. Also, the interval d between the microphones 9 in the example shown in FIG. 10 is the interval d at which a grating lobe due to external sound occurs within the maximum angle βrange of the orientation determination range TA when the delay-and-sum beamforming calculation is performed based on the sound pressure information acquired by the sound source orientation determination device 1.

[0040] In the example shown in FIG. 10, microphones 9 are arranged at regular intervals on the circumferences of two virtual circles that form concentric circles centered on the center of the array sensor 3. Specifically, among the two virtual circles, four microphones 9 are arranged at regular intervals on the outer virtual circle, and four microphones 9 are arranged at regular intervals on the inner virtual circle. Also, in the example shown in FIG. 10, the diameter of the outer virtual circle is approximately twice the diameter of the inner virtual circle among the two virtual circles. In the following description, it is assumed that the opening of the hood 8, that is, the inner diameter of the hood 8, is set to be sufficiently longer than the wavelength of the sound wave measured by the sound source direction calibration device 1. Also, it is assumed that the sound source is sufficiently far from the interval between the microphones 9 and a plane wave arrives at the array sensor 3 from the sound wave arrival direction indicated by the arrow. Further, applying the ray tracing method that does not particularly consider the diffraction phenomenon of the sound wave, the length of the hood 8 and the inner diameter of the hood 8 in the embodiment of the present invention will be described. In the example shown in FIG. 10, the interval between the microphone 9 and the hood 8 is the narrowest along the sound wave arrival direction. That is, FIG. 10 shows the sound wave arrival direction in which the microphone 9 closest to the hood 8 is most easily shielded by the hood 8.

[0041] FIG. 11 is a cross-sectional view taken along the line B - B shown in FIG. 10 for explaining the height of the hood 8 and the inner diameter of the hood 8 in the embodiment of the present invention. Here, let the angle formed by the line perpendicular to the surface of the array sensor 3 (hereinafter referred to as the normal line) and the peripheral edge OE of the direction calibration range TA be θmax. In the example shown in FIG. 11, the sound source direction calibration device 1 is configured to measure sound waves arriving from the front side. Also, the hood 8 is configured to shield sound waves from the peripheral edge OE of the direction calibration range TA for one of the plurality of microphones 9 installed on the array sensor 3. In FIG. 11, for the sound wave arriving from the angle θmax, the microphone 9-1 is shielded by the hood 8. Also, the sound wave arriving at the array sensor 3 from outside the angle θmax is the external sound in the embodiment of the present invention.

[0042] The symbol "a" shown in Fig. 11 indicates the distance between the microphone 9-1, which is located closest to the sound wave arrival direction among the plurality of microphones 9, and the hood 8. In the azimuth calibration of the sound source using the delay-and-sum beamforming calculation, it is necessary to receive the sound wave by at least two microphones 9. In the example shown in Fig. 11, when the sound wave arrives from the angle θmax, it is necessary to receive the sound wave by at least the microphones 9-3 and 9-4 on the side farthest from the hood 8 with respect to the arrival direction of the sound wave. That is, in the examples shown in Figs. 10 and 11, when the sound wave arrives from the angle θmax, if the sound wave can be received by the microphone 9-3, the sound wave can be received by at least two microphones 9-3 and 9-4. The symbol "ax" shown in Fig. 11 indicates the distance between the microphone 9-3 and the hood 8 in the sound wave arrival direction. When the sound wave arrives from the peripheral edge OE of the azimuth calibration range TA (from the angle θmax in Fig. 11), the height of the hood 8 that shields the sound wave from at least one microphone (the microphone 9-1 in Figs. 10 and 11) and enables the sound wave to be received by two or more microphones 9 (the microphones 9-3 and 9-4 in Figs. 10 and 11) can be expressed by the following formula (2). a·tan(90° - θmax) ≤ height of the hood < ax·tan(90° - θmax) ···(2)

[0043] FIG. 12 is a diagram for explaining the range shielded by the hood 8 when the height of the hood 8 is set to the minimum height that satisfies formula (2). On the array sensor 3 shown in FIG. 12, the area where the sound waves arriving from the peripheral edge OE of the azimuth calibration range TA are shielded by the hood 8 is hatched. Note that the arrangement of the microphones 9 in FIG. 12 is the same as the arrangement of the microphones 9 shown in FIG. 3 based on the above-described embodiment. When the height of the hood 8 is set so as to satisfy the lower limit value of the above formula (2), the hood 8 shields the sound waves arriving along the peripheral edge OE of the azimuth calibration range TA with respect to at least one microphone 9. Also, with respect to the remaining microphones 9 except for at least one microphone 9, the arrival of the sound waves is not shielded. Therefore, the remaining microphones 9 can receive the sound waves arriving along the peripheral edge OE of the azimuth calibration range TA.

[0044] Therefore, according to the embodiment of the present invention, by using the hood 8 shown in FIG. 12, although the sound wave signals arriving from the peripheral edge OE of the azimuth calibration range TA that can be received by the array sensor 3 decrease as a whole of the sound source azimuth calibration device 1, the reception of external noise can be suppressed. Therefore, when performing the delay-and-sum beamforming calculation, it is possible to suppress the generation of grating lobes within the azimuth calibration range TA. Also, as a whole of the sound source azimuth calibration device 1, it is possible to obtain sufficient signals required for azimuth calibration of the sound source, that is, sound pressure information from each azimuth within the azimuth calibration range TA. Therefore, even if there is a source of external noise outside the azimuth calibration range TA of the sound source azimuth calibration device 1, it is possible to prevent or suppress misjudgment of the sound source. As a result, the accuracy of azimuth calibration of the sound source can be improved.

[0045] FIG. 13 is a diagram for explaining a range shielded by the hood 8 when the height of the hood 8 is set to the maximum height that satisfies Expression (2). Note that the area where sound waves arriving from the peripheral edge OE of the azimuth determination range TA are shielded by the hood 8 is hatched. In the example shown in FIG. 13, among the plurality of microphones 9, when performing delay-and-sum beamforming calculation, the height of the hood 8 is set so that at least two necessary microphones 9 receive sound waves arriving along the peripheral edge OE of the sound source azimuth determination range TA.

[0046] In the configuration shown in FIG. 13, compared with the configuration shown in FIG. 12, the number of microphones 9 whose sound waves arriving from the peripheral edge OE of the azimuth determination range TA are shielded by the hood 8 increases. As a result, the shielding effect of external noise is improved, and the generation of grating lobes can be more suppressed. Therefore, even if there is a source of external noise outside the azimuth determination range TA of the sound source azimuth determination device 1, misjudgment of the sound source can be prevented or suppressed. As a result, the accuracy of azimuth determination of the sound source can be improved.

[0047] Furthermore, the effect of reducing misjudgment of the azimuth determination of the sound source within the azimuth determination range TA by the hood 8 also varies depending on the arrangement of the microphones 9 of the array sensor 3. Assuming that the microphone 9 is omnidirectional, for example, taking the front of the azimuth determination device 1 as the measurement direction (0°), consider the case where sound waves arrive from an azimuth of 90° on either the left or right side (the horizontal direction of the array sensor 3) with respect to the measurement direction. In that case, if the arrangement of the microphones 9 is such that grating lobes are generated within the azimuth determination range TA of the sound source azimuth determination device 1, the hood 8 that shields the sound waves arriving from the azimuth on the one side (90°) is effective. However, in the case of the arrangement of the microphones 9 where no grating lobes are generated within the azimuth determination range TA, the hood 8 has no effect of preventing misjudgment by shielding sound sources outside the azimuth determination range TA.

[0048] When a sound source exists in front of the sound source azimuth calibration device 1, when delay-and-sum beamforming calculation is performed based on the acquired sound pressure information, a signal with a high sound pressure from the sound source, called the main lobe, appears at the position of 0° in front of the azimuth calibration device 1. The angle α (°) of the occurrence azimuth of the grating lobe with respect to the azimuth angle (which may be called the reference angle) where the main lobe appears can be expressed by the following formula (3). α = 90° / (d / λ) ···(3) The above d is the arithmetic mean value of the distances from each of the plurality of microphones 9 to the nearest microphone 9, and the above λ is the wavelength of the sound wave emitted from the sound source.

[0049] From the above formula (3), it can be understood that as the ratio (d / λ) of the interval d and the sound wave wavelength λ increases, the occurrence angle α of the grating lobe decreases. That is, when the interval d between the microphones 9 is increased, the occurrence angle α of the signal (not shown) indicating the grating lobe appears closer to the main lobe (0°).

[0050] Also, when sound waves arrive from a direction (90°) orthogonal to the front, which is the maximum sound wave arrival angle that can be incident on each microphone, with respect to the microphones 9, the condition for the grating lobe to occur within the maximum angle βrange of the azimuth calibration range TA can be expressed by the following formula (4). d ≧ 90°×λ / (90° + βrange / 2) ···(4) That is, when the interval d between the microphones 9 satisfies the formula (4), a signal indicating the grating lobe appears within the maximum angle βrange of the azimuth calibration range TA. Also, as described above, when the interval d increases, a signal indicating the gray lobe appears closer to the main lobe (0°) than when the interval d is small.

[0051] As described above, in the embodiment of the present invention, even when the distance d between the microphones 9 satisfies the formula (4) and is a distance d that generates a grating lobe due to external noise within the maximum angle βrange of the azimuth calibration range TA, the hood 8 is configured to physically shield part or all of the external noise from the array sensor 3. Therefore, when delay-and-sum beamforming calculation is performed based on the sound pressure information acquired by the sound source azimuth calibration device 1, it is possible to suppress the appearance of a grating lobe within the azimuth calibration range TA. As a result, misjudgment of the sound source can be suppressed.

[0052] Further, in the embodiment of the present invention, even if the distance d between the microphones 9 is set large so as to satisfy the formula (4), it is possible to suppress the generation of a grating lobe caused by external noise within the azimuth calibration range TA of the sound source. Therefore, even when the number of microphones 9 is small, the aperture width of the array sensor 3 can be increased to improve the directivity. As a result, the accuracy of sound source azimuth calibration can be improved without increasing the number of components, the calculation load, etc.

[0053] On the other hand, when delay-and-sum beamforming calculation is executed, side lobes are generated around the main lobe separately from the grating lobe. FIG. 14 is a diagram showing a part of the sound pressure map calculated by delay-and-sum beamforming as a one-dimensional sound wave distribution when there is a sound source in the front (0°) of the sound source azimuth calibration device 1. The vertical axis in FIG. 14 indicates the sound pressure (dB), and the horizontal axis indicates the angle for searching the sound source azimuth. According to FIG. 14, it can be seen that the main lobe appears in the front (0°) of the sound source azimuth calibration device 1, and side lobes are generated around the main lobe. It is generally known that the side lobes are reduced by applying a window function to the output of each sampling point. Therefore, by regarding each microphone 9 as each sampling point and combining the shielding effect by the hood 8 and the correction (weighting) of the detection sensitivity of the microphone 9 not shielded by the hood 8, it is possible to obtain the same effect as the reduction of the side lobes by the window function. Note that the detection sensitivity of the microphone 9 is the sensitivity with respect to the received sound pressure of the microphone 9.

[0054] In another embodiment of the present invention, weighting, i.e., correction, is applied to the detection sensitivity of the microphone 9 not shielded by the hood 8, and combined with the attenuation effect due to the diffraction of sound waves arriving from the peripheral edge OE of the azimuth calibration range TA by the hood 8. Thereby, it is configured to reduce the generation of side lobes with respect to the main lobe in the vicinity of the peripheral edge OE of the azimuth calibration range TA when performing delay-and-sum beamforming calculation. FIG. 15 is a flowchart for explaining an example of control executed in another embodiment of the present invention. The routine shown in FIG. 15 is repeatedly executed at predetermined short time intervals by the above-described arithmetic means 4.

[0055] First, the sound wave (n - 1) in the first (n - 1) measurement is acquired for each microphone 9 of the array sensor 3 (step S1), and band-pass filter processing is performed on each of the acquired sound waves (n - 1) (step S2). Here, n is a natural number of 2 or more. As a result, among the acquired sound waves (n - 1), the sound waves in the preset frequency band used for searching for the sound source are extracted, and the sound waves in the unnecessary frequency band are removed. Delay-and-sum beamforming calculation is performed on the sound waves extracted in step S2, and a sound pressure map (n - 1) corresponding to the sound wave (n - 1) acquired in step S1 is created (step S3).

[0056] In Fig. 16(a), an example of the received sound pressure distribution for each microphone 9 obtained in step S1 is connected by a smooth line and shown as a thick solid line. In the example shown in Fig. 16, there is a sound source (not shown) in the right direction of the sound source azimuth calibration device 1 (the right direction in Fig. 16), and sound waves are arriving from the sound source at the sound source azimuth calibration device 1. The sound waves incident on the microphones 9 located on the arrival direction side of the sound waves, that is, on the sound source side (the right side in Fig. 16), are subject to the shielding effect of the hood 8. Therefore, as shown by the thick solid line in Fig. 16(a), the sound pressure of the sound waves incident on the microphones 9 subject to the shielding effect of the hood 8 is lower than the sound pressure of the sound waves incident on the microphones 9 that are less subject to the shielding effect of the hood 8. Also, as the distance between the hood 8 and the microphones 9 on the arrival direction side of the sound waves increases, the sound pressure incident on the microphones 9 gradually increases. This is due to the diffraction phenomenon of sound waves.

[0057] Based on the above sound pressure map (n - 1), the azimuth of the sound source is estimated (step S4). Next, the sensitivity correction coefficient (n - 1) for each microphone 9 is determined (step S5). That is, there is a correlation between the estimated azimuth of the sound source and the received sound pressure of each microphone 9 affected by the shielding of the hood 8, and the relative sound pressure distribution for each microphone 9 according to the azimuth of the sound source can be estimated based on the dimensions of the hood 8 and the azimuth of the sound source. Based on the estimated sound pressure distribution, the correction coefficient for each microphone 9 is determined according to the position of the microphone 9 not affected by the shielding of the hood 8.

[0058] Subsequent to the process in step S5, the sound wave (n) in the second (n) measurement is acquired (step S6), and band - pass filter processing is performed on the sound wave (n) (step S7). The process in step S7 is the same as the process in step S2. Subsequently, the sound wave (n) extracted in step S7 is multiplied by the sensitivity correction coefficient (n - 1) determined in step S5 for correction (step S8).

[0059] Then, for each of the sound waves (n) corrected in step S8, delay-and-sum beamforming calculation is performed, and a sound pressure map (n) corresponding to the sound wave (n) acquired in step S6 is created (step S9). The sound pressure map (n) is output from the arithmetic means 4 to the display means 5 and displayed. Subsequently, based on the sound pressure map (n) created in step S9, the position of the sound source is determined (step S10). Note that the position of the sound source may be displayed on the display means 5.

[0060] In Fig. 16(b), an example of the received sound pressure distribution for each microphone 9 after sensitivity correction determined by the position of each microphone 9 and the degree of sound source shielding by the hood 8 is connected by a smooth line and shown by a dotted line. As shown by the dotted line in Fig. 16(b), by performing the correction in step S8, the detection sensitivity of the microphones 9 arranged at positions far from the sound source and not affected by the shielding of the hood 8 is reduced. The received sound pressure distribution for each microphone 9 shown by the dotted line in Fig. 16(b) has a similar shape to the sound pressure distribution after multiplying the window function for each microphone 9 when suppressing the generation of side lobes using the window function when not affected by the shielding of the hood 8.

[0061] Therefore, according to the embodiment of the present invention, by performing sensitivity correction on the detection sensitivity of the microphones 9 not shielded by the hood 8, the side lobes can be reduced in substantially the same manner as when the side lobes are corrected based on the window function. Therefore, even when there is a sound source in the vicinity of the peripheral edge OE of the azimuth calibration range TA and the intensity of the main lobe (sound source signal) in the peripheral edge OE of the azimuth calibration range TA decreases due to the use of the hood 8, the main lobe can be clearly discriminated. That is, a decrease in the azimuth calibration accuracy of the sound source azimuth by the sound source azimuth calibration device 1 can be suppressed.

[0062] Also, the sensitivity correction coefficient described above is changed in magnitude according to the size such as the height and inner diameter of the hood 8. This is because the range in which sound waves arriving from the peripheral edge OE of the azimuth calibration range TA are shielded and diffracted changes depending on the size of the hood 8. FIG. 17 is a diagram showing an example of the received sound pressure distribution for each microphone 9 when the height of the hood 8 is higher than that of the hood 8 shown in FIG. 16. In the example shown in FIG. 17, the area shielded by the hood 8 is larger than that in the example shown in FIG. 16. Also, as shown by the thick solid line in FIG. 17(a), the microphones 9 whose received sound pressure is reduced by the hood 8 reach closer to the center of the array sensor 3 compared to FIG. 16(a).

[0063] Then, based on the received sound pressure distribution shown in FIG. 17(a), the processes of steps S4 to S8 shown in FIG. 15 are executed to determine the sensitivity correction coefficient for each microphone 9, and the detection sensitivity for each microphone 9 is corrected by the sensitivity correction coefficient. The received sound pressure distribution of each microphone 9 including the shielding effect by the hood 8 after correcting the detection sensitivity for each microphone 9 in this way is shown in FIG. 17(b). The received sound pressure distribution for each microphone 9 shown in FIG. 17(b) has a shape similar to the sound pressure distribution after multiplying the window function for each microphone 9 when suppressing the generation of side lobes using the window function. Therefore, even when the height of the hood 8 is high, almost the same operations and effects as those of the embodiment shown in FIG. 16 can be obtained.

[0064] (Example) Next, an example conducted to confirm the effect of the hood 8 in the embodiment of the present invention will be described. Microphones 9 were arranged at regular intervals on the circumferences of two virtual circles that are concentric circles with different outer diameters from each other to form the array sensor 3. Specifically, the outer diameter of the inner virtual circle in the radial direction was set to 15.3 mm, and six microphones 9 were arranged at regular intervals on the circumference of the virtual circle. The outer diameter of the outer virtual circle in the radial direction, that is, the aperture width of the array sensor 3, was set to 41.0 mm, and seven microphones 9 were arranged at regular intervals on the circumference of the virtual circle. The arithmetic mean value of the intervals d between adjacent microphones 9 was 10.8 mm, the wavelength of the sound source, that is, the measured wavelength λ at the array sensor 3 was 8.7 mm, and the ratio (d / λ) of the interval d to the measured wavelength λ was 1.2. Also, the maximum angle βrange of the sound source orientation calibration range TA in the sound source orientation calibration device 1 was set to 62°. Further, a square cylinder with an inner dimension of 60 mm × 60 mm and a height of 43 mm was used as the hood 8. Since the other configurations are the same as those shown in FIGS. 1 to 3, the description thereof is omitted. Note that the height of the hood 8 satisfies the above-described formula (2).

[0065] (Experimental method) The sound source azimuth calibration device 1 configured as described above was installed on a predetermined placement surface so as to rotate about a rotation center axis (not shown) extending in the vertical direction and to maintain a substantially horizontal orientation. With the front of the sound source azimuth calibration device 1 as the reference position (0°), a sound source to be azimuth calibrated (hereinafter referred to as the first sound source) was installed at that reference position. Also, an external sound source (hereinafter referred to as the second sound source) was installed at a position 50° in either the left or right direction with respect to the reference position when viewed from the sound source azimuth calibration device 1. Then, the sound source azimuth calibration device 1 was rotated about the rotation center axis by 50° increments from the reference position of the sound source azimuth calibration device 1 to calibrate the azimuths of the first sound source and the second sound source. Also, the sound pressure emitted by the first sound source was kept constant, and the sound pressure of the external sound emitted by the second sound source was increased relative to the sound pressure emitted by the first sound source. By doing so, it was evaluated whether the azimuth of the first sound source could be calibrated even when external sound with a higher sound pressure than the sound wave emitted by the first sound source reached the sound source azimuth calibration device 1. Also, the azimuth calibration of the first sound source was performed multiple times using such a sound source azimuth calibration device 1, and when the ratio of the actual installation position of the first sound source to the azimuth calibration result was 50% or more, it was determined that the azimuth of the first sound source could be calibrated.

[0066] (Comparative Example 1) It was configured in the same manner as in the example except that the hood 8 was not provided.

[0067] (Comparative Example 2) It was configured in the same manner as in the example except that the height of the hood 8 was set lower than the lower limit height of the hood 8 shown by the above formula (2).

[0068] (Evaluation) In Comparative Example 1, when the sound pressure of the second sound source (external sound) was up to 2.3 times that of the first sound source, the azimuth calibration of the first sound source was possible. In Comparative Example 2, when the sound pressure of the second sound source (external sound) was up to 4.0 times that of the first sound source, the azimuth calibration of the first sound source was possible. On the other hand, in the embodiment of the present invention, when the sound pressure of the second sound source (external sound) was up to 8.2 times that of the first sound source, the azimuth calibration of the first sound source was possible. Thus, by using the hood 8 in the embodiment of the present invention, it was confirmed that even when a strong external sound source exists outside the azimuth calibration range TA of the sound source azimuth calibration device 1, the azimuth calibration of the sound source within the azimuth calibration range TA is possible.

Explanation of Reference Numerals

[0069] 1 Sound source azimuth calibration device 2 Camera 3 Array sensor 4 Arithmetic means 5 Display means 6 Input means 7 Housing 8 Hood 9 Microphone 10 Processing unit 11 Storage unit TA Sound source azimuth calibration range in the sound source azimuth calibration device βrange Maximum angle of the sound source azimuth calibration range in the sound source azimuth calibration device

Claims

A sound source direction calibration device comprising a sound wave shielding hood configured to suppress reception of external sound arriving from outside a preset direction calibration range by an array sensor having a plurality of microphones for receiving sound waves, wherein the sound wave shielding hood is configured to shield sound waves arriving from the peripheral portion of the direction calibration range for at least one of the plurality of microphones, and comprising arithmetic means for calculating the sound pressure of each direction within the direction calibration range based on the sound pressure information of the sound waves received by the array sensor, and calibrating, as the arrival direction of the sound waves from the sound source, the direction at which the sound pressure is maximum among the respective directions, A sound source direction calibration device comprising a sound wave shielding hood, wherein an average distance d between adjacent ones of the plurality of microphones satisfies the following formula: d ≧ 90° × λ / (90° + βrange / 2) (mm) where λ above is the center wavelength (mm) of the sound wave used for calculation of the arrival direction of the sound wave, and βrange is the maximum angle (°) of the direction calibration range. A sound source direction calibration device comprising a sound wave shielding hood configured to suppress reception of external sound arriving from outside a preset direction calibration range by an array sensor having a plurality of microphones for receiving sound waves, wherein the sound wave shielding hood is configured to shield sound waves arriving from the peripheral portion of the direction calibration range for at least one of the plurality of microphones, and comprising arithmetic means for calculating the sound pressure of each direction within the direction calibration range based on the sound pressure information of the sound waves received by the array sensor, and calibrating, as the arrival direction of the sound waves from the sound source, the direction at which the sound pressure is maximum among the respective directions, wherein the arithmetic means is configured to weight the sound pressure information of the sound waves received by each of the plurality of microphones according to the height of the sound wave shielding hood, and calibrate the direction of the sound source based on the weighted sound pressure information. A sound source direction calibration device comprising a sound wave shielding hood.

Images