Sound source location identification method and sound source location identification device

Abstract

Even if the sound emitted from the sound source is of high frequency, the location of the sound source can be determined with high precision. [Solution] The sound source location device 3 includes a controller 21 that identifies the location of a sound source 13 of a sound-emitting object 11. The controller 21 acquires sound signals from the sound source 13 detected by a microphone array 5 equipped with multiple microphones Ma to Mn, and selects the microphone with the highest quality detected sound signal among the multiple microphones Ma to Mn as the first microphone. The controller 21 calculates a transfer function between the first sound signal detected by the first microphone and the second sound signal detected by microphones other than the first microphone, corrects the calculated transfer function to reduce noise components, and corrects the second sound signal using the corrected transfer function and the first sound signal. The controller 21 identifies the location of the sound source using the corrected second sound signal and the first sound signal.

Description

[Technical Field]

[0001] This invention relates to a method for determining the location of a sound source and a device for determining the location of a sound source. [Background technology]

[0002] Conventionally, a beamforming technique usable for sound source detection has been disclosed in Patent Document 1. In the nonlinear beamforming disclosed in Patent Document 1, a filter was designed that forms directional characteristics by arranging multiple microphone arrays in an arbitrary space and combining them through the output signals of multiple microphones. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2006-270903 [Overview of the project] [Problems that the invention aims to solve]

[0004] When the sound emitted from a sound source is of a high frequency, the directivity of the sound increases. In the conventional sound source detection method using beamforming described above, only a specific microphone among multiple microphones receives a signal with a high signal-to-noise ratio, while the other microphones receive a relatively low signal-to-noise ratio. Therefore, there was a problem in that the location of the sound source could not be determined with high precision.

[0005] This invention has been made in view of the above problems, and its purpose is to provide a sound source location method and apparatus that can pinpoint the location of a sound source with high accuracy, even when the sound emitted from the sound source is of a high frequency. [Means for solving the problem]

[0006] The sound source position identification method and its apparatus according to one aspect of the present invention select, as the first microphone, the microphone with the highest detected sound signal quality among a plurality of microphones, calculate the transfer function between the first sound signal detected by the first microphone and the second sound signal detected by microphones other than the first microphone, correct the calculated transfer function so that the noise component is reduced, correct the second sound signal using the corrected transfer function and the first sound signal, and identify the position of the sound source using the corrected second sound signal and the first sound signal.

Effect of the Invention

[0007] According to one aspect of the present invention, even if the sound emitted from the sound source is a high frequency, the position of the sound source can be identified with high accuracy.

Brief Description of the Drawings

[0008] [Figure 1] FIG. 1 is a block diagram showing the configuration of a sound source position identification system according to an embodiment. [Figure 2] FIG. 2 is a flowchart showing a sound source position identification method according to an embodiment.

Modes for Carrying Out the Invention

[0009] Hereinafter, referring to the drawings, the sound source position identification method and the sound source position identification apparatus according to the present embodiment will be described. In the description of the drawings, the same reference numerals are given to the same parts and detailed description thereof will be omitted.

[0010] Referring to FIG. 1, the configuration of the sound source position identification system according to the present embodiment will be described. As shown in FIG. 1, the sound source position identification system 1 includes a sound source position identification apparatus 3, a microphone array 5, a camera 7, and a display device 9.

[0011] The microphone array 5 is a sound collection device including a plurality of microphones Ma to Mn. For example, in the microphone array 5, a plurality of microphones Ma to Mn are provided at predetermined intervals in an annular housing, and it is arranged facing a measurement target object 11 such as an engine or a motor. Each of the microphones Ma to Mn detects the sound pressure of the sound emitted from the sound source 13 of the measurement target object 11, and outputs a sound signal corresponding to the detected sound pressure to the sound source position identification device 3.

[0012] The camera 7 is arranged at the center of the annular housing of the microphone array 5, and outputs an image obtained by imaging the measurement target object 11 to the sound source position identification device 3.

[0013] The display device 9 is a display that displays the position of the identified sound source 13 on the image captured by the camera 7.

[0014] The sound source position identification device 3 is a device that identifies the position of the sound source 13 of the measurement target object 11 that emits sound. Specifically, the sound source position identification device 3 performs a sound source search for identifying the position of the sound source 13 that emits a sound such as abnormal noise from among the measurement target objects 11 such as an engine or a motor by using the sound signals detected by the plurality of microphones Ma to Mn.

[0015] In particular, when the sound emitted from the sound source 13 is a high frequency, the directivity of the sound becomes high, so the number of microphones that can receive a signal with a high S / N ratio is limited. Therefore, only a specific microphone detects a signal with a high S / N ratio, and the S / N ratio of many other microphones decreases. Therefore, the sound source position identification device 3 performs a sound source search so that the position of the sound source 13 can be accurately identified even when the sound emitted from the sound source 13 is a high frequency. The sound source position identification device 3 includes a controller 21 and a storage device 23.

[0016] The controller 21 performs a process to identify the location of the sound source 13 of the object 11 that emits sound. Specifically, the controller 21 acquires the sound signal from the sound source 13 detected by the microphone array 5 and selects the microphone with the highest quality detected sound signal among multiple microphones Ma to Mn as the first microphone. The controller 21 then calculates the transfer function between the first sound signal detected by the first microphone and the second sound signal detected by the microphones other than the first microphone, and corrects the calculated transfer function so that the noise component is reduced. Furthermore, the controller 21 corrects the second sound signal using the corrected transfer function and the first sound signal, and uses the corrected second sound signal and the first sound signal to identify the location of the sound source 13.

[0017] The controller 21 consists of a hardware processor, various types of memory, input / output interfaces, etc. The hardware processor may be, for example, a CPU (Central Processing Unit) or an MPU (Micro-Processing Unit). Various programs are installed on the controller 21, and the controller 21 can perform various processes by executing these programs.

[0018] The storage device 23 is a memory or database that stores information necessary to identify the location of the sound source 13. Specifically, the storage device 23 stores sound signals acquired from the microphone array 5, and stores multiple sound signals detected by multiple microphones Ma to Mn. The storage device 23 also stores images acquired from the camera 7.

[0019] Next, a method for determining the location of a sound source using the sound source location device 3 according to this embodiment will be described. Figure 2 is a flowchart showing the processing steps of the sound source location determination process. As shown in Figure 2, in step S10, the controller 21 acquires an audio signal from the sound source 13 detected by the microphone array 5 equipped with multiple microphones Ma to Mn.

[0020] For example, the sound signal A from microphone Ma is expressed in the frequency domain by equation (1) below, and the sound signal B from microphone Mb is expressed in the frequency domain by equation (2) below. The controller 21 acquires such sound signals from all microphones. Here, S is the sound signal from sound source 13, Zb is the noise component, and Ha and Hb are the transfer functions of microphones Ma and Mb.

number

number

[0021] In step S11, the controller 21 selects the microphone with the highest quality detected sound signal from among the multiple microphones Ma to Mn as the first microphone. For example, the controller 21 selects the microphone that detected the sound signal with the maximum sound pressure at a specific frequency to be measured as the first microphone, as it is the microphone with the highest quality.

[0022] Alternatively, the microphone with the highest signal-to-noise ratio of the detected sound signal may be selected as the first microphone, as it represents the highest quality microphone. For example, in the case of microphones Ma and Mb, as shown in equation (1), the sound signal A from microphone Ma does not contain any noise components, so the quality of sound signal A is the highest, and microphone Ma is selected as the first microphone.

[0023] In step S12, the controller 21 calculates the transfer function between the first sound signal detected by the first microphone and the second sound signal detected by the microphones other than the first microphone. Hereinafter, the first microphone will be referred to as microphone Ma (or "first microphone Ma" as needed), and the microphones other than the first microphone will be referred to as microphone Mb.

[0024] First, represent the second sound signal detected by the microphone Mb using the first sound signal detected by the first microphone Ma. For example, the sound signal B, which is the second sound signal, can be expressed as the transfer function H * , AB , AB , AN , the sound signal A, which is the first sound signal, and the noise component Z.

Equation

[0025] Then, multiply both sides of Equation (3) by the complex conjugate A * of the sound signal A to transform the equation as shown in Equation (4). As a result, the transfer function H AB between the first sound signal A and the second sound signal B is calculated.

Equation

[0026] Similarly, for each of the microphones Mc to Mn other than the microphone Mb, the transfer function H AC to H AN between the first sound signal A detected by the first microphone and the sound signals (second sound signals) C to N detected by the microphones Mc to Mn is calculated.

[0027] In this way, the transfer functions H<00^00006> to H AN of the microphones Mb to Mn other than the first microphone Ma, that is, the microphones Mb to Mn with low quality, are expressed using the sound signal (first sound signal) A of the first microphone Ma, which has the highest quality, the complex conjugate A * of the sound signal A, and the sound signals (second sound signals) B to N of the microphones Mb to Mn.

[0028] In step S13, the controller 21 corrects the transfer function calculated in step S12 so that the noise component is reduced. For example, the transfer function H ABIn response, the controller 21 corrects the noise component by the averaging process shown in equation (5). For example, the controller 21 divides the time-domain sound signals A and B into multiple segments (N: number of segments) and applies a transfer function H to each of them. AB Calculate and add them together.

number

[0029] The noise component ZA shown in equation (5) * While has the characteristic of appearing randomly, BA * This is the cross-spectrum of sound signals A and B, and has certain characteristics. When these components are added together segment by segment, BA * Noise component ZA * The value becomes relatively small. Therefore, by performing the averaging process shown in equation (5), the noise component can be reduced and the S / N ratio can be improved. The controller 21 has a transfer function H AC ~H AN Similarly, averaging is performed on the data to correct it and reduce the noise component.

[0030] In step S14, the controller 21 corrects the second tone signal using the transfer function corrected in step S13 and the first tone signal. Specifically, by representing the second tone signal B with the transfer function corrected in step S13 and the first tone signal A, the corrected tone signal B is obtained as shown in equation (6). new Similarly, the corrected sound signal C is required. new ~N new This is required.

number

[0031] Thus, the audio signal B from a microphone of low quality Mb~Mn new ~N new This is a transfer function H with reduced noise components. ABThis is expressed as the product of the highest quality microphone Ma and the sound signal A. As shown in equation (1), the sound signal A has a high A / N ratio such that the noise component can be ignored. Therefore, as shown in equation (2), the corrected sound signal B is used to pinpoint the location of the sound source 13, rather than using the sound signal B which contains the noise component Zb. new ~N new Using this method to pinpoint the location of sound source 13 allows for more accurate pinpointing of the sound source 13's location.

[0032] In step S15, the controller 21 receives the sound signal (first sound signal) A and the sound signal (second sound signal) B corrected in step S14. new ~N new The location of sound source 13 is determined using this method.

[0033] For example, the controller 21 receives sound signal A and corrected sound signal B. new ~N new The location of the sound source 13 is determined using the following method. Specifically, a general beamforming sound source detection technique can be used to determine the location of the sound source 13. In this sound source detection technique, the sound pressure distribution is detected using phase difference information from the sound source 13 to microphones Ma to Mn to determine the location of the sound source 13. Once the location of the sound source 13 is determined in this way, the controller 21 displays the location of the sound source 13 on the image of the camera 7 displayed on the display device 9, and the sound source location determination process according to this embodiment is completed.

[0034] As described in detail above, the sound source location identification device 3 according to this embodiment selects the microphone with the highest quality detected sound signal from among a plurality of microphones as the first microphone, calculates a transfer function between the first sound signal detected by the first microphone and the second sound signal detected by microphones other than the first microphone, corrects the calculated transfer function so as to reduce the noise component, corrects the second sound signal using the corrected transfer function and the first sound signal, and identifies the location of the sound source using the corrected second sound signal and the first sound signal.

[0035] This results in the transfer function H of a microphone Mb~Mn of low quality. AB ~H AN The audio signal A of the highest quality microphone Ma is used to calculate the corrected audio signal B. Furthermore, the audio signals B through N of the lower quality microphones Mb through Mn are corrected by the audio signal A of the highest quality microphone Ma. Therefore, the corrected audio signal B is calculated from the audio signal A of the highest quality microphone Ma. new ~N new This can be represented by the sound signal A from the highest quality microphone Ma, and this corrected sound signal B new ~N new Since the position of the sound source 13 is determined using this method, the position of the sound source 13 can be determined with high accuracy even if the sound emitted from the sound source 13 is of a high frequency.

[0036] Furthermore, the sound source location device 3 according to this embodiment corrects the transfer function by averaging to reduce noise components. This makes it easy to reduce noise components in the transfer function, allowing the location of the sound source 13 to be determined with high accuracy.

[0037] As described above, embodiments of the present invention have been presented, but the statements and drawings that constitute part of this disclosure should not be understood as limiting the invention. Various alternative embodiments, examples, and operational techniques will become apparent to those skilled in the art from this disclosure. [Explanation of Symbols]

[0038] 1: Sound source location system, 3: Sound source location device, 5: Microphone array, 7: Camera, 9: Display device, 11: Object to be measured, 13: Sound source, 21: Controller, 23: Memory device, A~N: Sound signal, Ma~Mn: Microphone

Claims

1. A method for determining the location of a sound source of an object to be measured that emits sound, which is performed by a controller that determines the location of the sound source, The sound signal from the sound source detected by the microphone array, which has multiple microphones, is acquired. Among the multiple microphones, the microphone with the highest quality detected sound signal is selected as the first microphone. The transfer function between the first sound signal detected by the first microphone and the second sound signal detected by a microphone other than the first microphone is calculated. The calculated transfer function is corrected to reduce the noise component. The second sound signal is corrected using the corrected transfer function and the first sound signal. The position of the sound source is determined using the corrected second sound signal and the first sound signal. Sound source localization method.

2. The transfer function is corrected by averaging to reduce noise components. The method for determining the location of a sound source according to claim 1.

3. A sound source location device equipped with a controller that identifies the location of the sound source of an object to be measured that emits sound, The aforementioned controller, The sound signal from the sound source detected by the microphone array, which has multiple microphones, is acquired. Among the multiple microphones, the microphone with the highest quality detected sound signal is selected as the first microphone. The transfer function between the first sound signal detected by the first microphone and the second sound signal detected by a microphone other than the first microphone is calculated. The calculated transfer function is corrected to reduce the noise component. The second sound signal is corrected using the corrected transfer function and the first sound signal. The position of the sound source is determined using the corrected second sound signal and the first sound signal. Sound source localization device.

Images