KR101483513B1 - Apparatus for sound source localizatioin and method for the same - Google Patents

Abstract

The present invention relates to a sound source position tracking apparatus and a sound source position tracking method, and more particularly, a sound source position tracking apparatus according to an embodiment of the present invention includes a plurality of microphones for sensing an applied sound wave to generate analog sound signals, And a sound processing unit for generating a digital sound signal corresponding to a change in size of a waveform of the analog sound signal and calculating a time delay of a sound wave applied to each microphone using the digital sound signal, It is possible to generate a digital sound signal by sampling the sound signal to generate sampling data, and to compare sizes between adjacent sampling data.

Description

FIELD OF THE INVENTION [0001] The present invention relates to a sound source location tracking device and a sound source location tracking method,

The present invention relates to a sound source position tracking device and a sound source position tracking method, and more particularly, to a sound source position tracking device and a sound source position tracking method capable of quickly tracking a sound source position using a digital sound signal representing the appearance of a sound wave.

The conventional sound source location tracking apparatus converts the input analog sound signal into digital data having a width of at least a byte and performs sound source location tracking using the digital data. For example, when the output of the analog-to-digital converter AD of the sound source location tracking device is 12 bits, the analog sound signal input to the microphone of the sound source location tracking device after the sound wave is generated from the sound source, And converted to 12-bit digital data, and the subsequent processing for tracking the sound source based on the digital data is performed.

In an ideal case, the amount of calculation for sound source position tracking is not a problem, but in actuality, sound waves input to a microphone may include noise due to echo and noise due to environmental characteristics such as propagation of sound waves. In this case, various additional algorithms for correcting this will be included, and the application of this additional algorithm will require a relatively long signal processing time. Therefore, real-time processing of the input analog acoustic signal becomes difficult, and in order to overcome this, demands for more advanced hardware specifications are required, resulting in an increase in the cost of the entire system.

Korean Patent Publication No. 10-2009-0038697 (October 16, 2007)

The present invention provides a sound source position tracking device and a sound source position tracking method that can quickly track a sound source position using a digital sound signal representing the external shape of a sound wave.

According to an embodiment of the present invention, there is provided an apparatus for tracking a sound source, comprising: a plurality of microphones for sensing an applied sound wave and generating an analog sound signal, respectively; And an acoustic processor for generating a digital acoustic signal corresponding to a change in magnitude of a waveform of the analog acoustic signal and calculating a time delay of a sound wave applied to each of the microphones using the digital acoustic signal, The processing unit may generate the digital sound signal by sampling the analog sound signal to generate sampling data, and comparing sizes between adjacent sampling data.

delete

If the i-th sampled data is greater than or equal to the (i-1) th sampled data, the acoustical processor outputs 1. If the i-th sampled data is less than or equal to the So as to generate the digital sound signal. At this time, i may be a natural number of 2 or more.

Here, the digital sound signal may be a binary sequence represented by one bit.

Here,

y _m is a digital sound signal, x _m is sampling data, and i is a natural number of 2 or more, thereby generating a 1-bit digital sound signal composed of 0 and 1.

Here, the sound processor calculates a cross-correlation between each digital sound signal using GCC (Generalized Cross-Correlation), and calculates a cross-correlation value between the digital sound signals using the cross- The corresponding time delay can be calculated.

Here,

는 시간지연,

는 디지털음향신호의 상호상관값, y_m은 디지털음향신호, x_m은 샘플링 데이터, i는 2 이상의 자연수, N은 샘플링 데이터의 개수를 이용하여, 상기 디지털음향신호에 대응하는 시간지연을 계산할 수 있다.

Time delay,

_Ym is a digital acoustic signal, _xm is sampling data, i is a natural number of 2 or more, and N is the number of sampling data, the time delay corresponding to the digital acoustic signal can be calculated have.

Wherein the sound processing unit samples the analog sound signal to generate sampling data, and outputs 1 if the sampling data to be input i is larger than i-1 (i is 2 or more natural number) input sampling data, An audio signal input unit for outputting 0 when the input sampling data is smaller than or equal to (i-1) th input sampling data, and generating a 1-bit digital audio signal corresponding to each of the analog audio signals; And applying a generalized cross-correlation (GCC) to each of the digital acoustic signals to calculate a cross-correlation between the digital acoustic signals, and using the cross- And a signal processor for calculating a corresponding time delay.

According to another aspect of the present invention, there is provided a sound source position tracking method, comprising: a sound wave sensing step of sensing an acoustic wave applied to a plurality of microphones when an acoustic wave is applied to generate an analog acoustic signal; A waveform extracting step of the acoustic processing unit generating a digital acoustic signal corresponding to the waveform of the analog acoustic signal; Calculating a time delay of a sound wave applied to the microphone using the digital sound signal; And a position determining step of determining a position of a sound source using the time delay, wherein the waveform extracting step comprises sampling the analog sound signal to generate sampling data, and comparing sizes of adjacent sampling data And generating the digital acoustic signal.

In addition, the means for solving the above-mentioned problems are not all enumerating the features of the present invention. The various features of the present invention and the advantages and effects thereof will be more fully understood by reference to the following specific embodiments.

According to the sound source location tracking device and the sound source location tracking method according to an embodiment of the present invention, data size and processing speed for sound source location tracking can be drastically reduced. In addition, since the data size and processing speed are reduced, the implementation cost of a complicated algorithm can be reduced, and the processing speed can be further improved.

1 is a block diagram illustrating an apparatus for tracking a sound source according to an embodiment of the present invention.
2 is a schematic diagram for explaining generation of a digital acoustic signal according to an embodiment of the present invention.
3 is a schematic diagram illustrating experimental conditions for testing the performance of a sound source location tracking apparatus according to an embodiment of the present invention.
4 is a schematic diagram and a table showing geometrical positions of a microphone and a sound source for testing a sound source performance according to an exemplary embodiment of the present invention.
FIG. 5 is a comparison chart comparing a time delay according to the number of samples measured by the sound source location tracking device according to an embodiment of the present invention and a time delay measured with a conventional sound source location tracking device.
FIG. 6 is a comparison chart comparing time delays according to the signal-to-noise ratio of the sound source location tracking apparatus according to an embodiment of the present invention and time delays according to the signal-to-noise ratio in the conventional sound source location tracking apparatus.
FIG. 7 is a comparison chart comparing a time delay between each sound source position and a time delay measured by a conventional sound source position tracking apparatus in the sound source position tracking apparatus according to an embodiment of the present invention.
8 is a flowchart illustrating a sound source position tracking method according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings, in order that those skilled in the art can easily carry out the present invention. In the following detailed description of the preferred embodiments of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear. In the drawings, like reference numerals are used throughout the drawings.

In addition, in the entire specification, when a part is referred to as being 'connected' to another part, it may be referred to as 'indirectly connected' not only with 'directly connected' . Also, to "include" an element means that it may include other elements, rather than excluding other elements, unless specifically stated otherwise.

1 is a block diagram illustrating an apparatus for tracking a sound source according to an embodiment of the present invention.

Referring to FIG. 1, the sound source location tracking apparatus according to an embodiment of the present invention may include a plurality of microphones 10 and an acoustic processor 20.

Hereinafter, a sound source position tracking apparatus according to an embodiment of the present invention will be described with reference to FIG.

The microphone 10 can sense an applied sound wave and generate an analog sound signal. The microphone 10 can sense the vibration of air caused by a sound wave generated from the sound source s and can generate an electrical signal corresponding to the vibration of the air. Here, the electrical signal converted by the sound wave is referred to as an analog sound signal.

As shown in FIG. 1, the microphone 10 may include a plurality of microphones including a first microphone 11, a second microphone 12, and the like. In this way, when a plurality of microphones 11, 12, ..., 1n are included, the time at which the sound waves arrive at the respective microphones 11 and 12 differs according to the position of the sound source s . That is, since the time at which each microphone 11, 12 starts to generate an analog acoustic signal corresponding to the sound wave is varied, a time delay may occur between the analog acoustic signals. In particular, as the microphones 11 and 12 are located far from the sound source s, the time delay may be larger.

Since the time delay occurs in proportion to the distance to the sound source s, the time delay between the analog sound signals generated by the respective microphones 10 is used to position the sound source s It is possible to discriminate. For example, since the speed of the sound wave is about 340 m / s, the distance between the microphones 11, 12 and 13 can be grasped from the sound source by using the time delay, 13) can be regarded as a known value. Therefore, the position of the sound source can be grasped by utilizing a trigonometric function or the like.

Specifically, the microphone phon 10 senses the sound wave to generate the analog sound signal, and then transmits the analog sound signal to the sound processing unit 20, and the sound processing unit 20 adjusts the time delay Can be calculated.

The acoustic processing unit 20 generates a digital acoustic signal corresponding to a change in the magnitude of the waveform of the analog acoustic signal and calculates the time delay of the acoustic wave applied to the microphone 10 using the digital acoustic signal . Here, the acoustic processor 20 may generate the sampling data by sampling the analog acoustic signal, and may generate the digital acoustic signal by comparing sizes between adjacent sampling data.

Conventionally, for example, sampling was performed 100 times per second to convert a value corresponding to the size of the analog sound signal into 12-bit digital data, and then the time delay was calculated using the converted digital data. As the number of sampling per second increases and the number of bits of the digital data increases, the digital data becomes closer to the analog sound signal. In contrast, the amount of digital data generated increases and the time of the sound wave The time required to calculate the delay increases. In particular, when the sound wave is influenced by environmental characteristics (echoes, reflection characteristics, etc.) in which the sound waves propagate, an additional algorithm or the like for correcting the noise is applied. When the calculation according to the application of the algorithm is performed, A long signal processing time may be required, and real-time processing may become difficult. In addition, in order to perform real-time processing, hardware specifications become higher, which may result in an increase in the cost required to construct the entire sound source location tracking device.

On the other hand, according to the acoustic processing unit 20 according to the embodiment of the present invention, the size of sampling data adjacent to each other can be compared to generate a 1-bit digital sound signal. By using the 1-bit digital sound signal in this manner, the amount of calculation for calculating the time delay of the sound wave can be drastically reduced, and quick result derivation can be performed as compared with the case of using the conventional 12-bit.

 Regarding the calculation of the time delay of the sound wave using the sound processing unit 20, the appearance of the signal appearing in the analog sound signal corresponds to an essential element in the calculation of the delay time. When sound waves are generated in the sound source s, sound waves of the same type are input to the first microphone 11 and the second microphone 12 with a predetermined time delay, respectively. Accordingly, it is possible to obtain the time delay of the sound waves by confirming whether the appearance of the signals represented by the respective analog sound signals generated by the microphones 11 and 12 are repeated in the same form with a certain time delay. Therefore, in order to calculate the time delay of the sound wave, not only the size value of the analog sound signal but also the change of the appearance of the signal of the analog sound signal is an important factor. Therefore, when the size of the sampling data adjacent to each other is compared with each other to generate the digital acoustic signal as in the acoustic processing unit 20 according to an embodiment of the present invention, the size of the analog acoustic signal unnecessary for the time delay calculation is excluded And information on the appearance of the analog acoustic signal can be obtained.

Specifically, the sound processing unit 20 may generate the digital sound signal in such a manner that 1 is output if the i-th sampled data is larger than the (i-1) -th sampled data, and 0 otherwise. At this time, i is a natural number of 2 or more. That is, whether the analog acoustic signal is increased or decreased can be represented by the digital acoustic signal, thereby obtaining information on the external appearance of the analog acoustic signal. In this case, the digital acoustic signal may be represented by a 1-bit binary sequence expressed by 0 or 1.

FIG. 2 (a) is a view showing sampling of an analog acoustic signal according to a conventional method and then obtaining corresponding digital data. FIG. 2 (b) And then expressed as a digital acoustic signal indicating whether the analog acoustic signal is increasing or decreasing. As shown in FIG. 2 (a), when each sampled sampling data itself is represented by 12-bit digital data, the amount of data to be processed in the calculation of the time delay thereafter increases, Information will be included in many.

On the other hand, as shown in FIG. 2 (b), if the information about whether the analog sound signal increases or decreases is expressed as a digital sound signal represented by 0 and 1, it is represented by a binary string of only 1 bit. And the computation time can be minimized.

Specifically, the sound processing unit 20 includes:

y _m is a digital sound signal, x _m is sampling data, and i is a natural number of 2 or more. x ₁ denotes sampling data for the analog sound signal received from the first microphone 11 and x ₂ denotes sampling data for the analog sound signal received from the second microphone 12.

Thereafter, the sound processing unit 20 may apply GCC (Generalized Cross-Correlation), which is a method for obtaining a time delay for the digital sound signal, by the GCC, The correlation value can be obtained. The GCC is a technique for determining the consistency between two signals by receiving two arbitrary signals and is one of widely used techniques for estimating a time delay in relation to sound source location tracking. The GCC is applied to the 1-bit digital sound signal. In this case, the GCC is applied to the 1-bit digital sound signal so that the time delay corresponding to each of the microphones 11 and 12 is Can be calculated.

Specifically, the sound processing unit 20

는 시간지연,

는 디지털음향신호의 상호상관값, y_m은 디지털음향신호, x_m은 샘플링 데이터, i는 2 이상의 자연수, N은 샘플링 데이터의 개수를 이용하여 상기 디지털음향신호에 대응하는 시간지연을 계산할 수 있다. 특히, 여기서는 1bit의 디지털음향신호를 활용하므로, 종래와 달리 XNOR를 활용하여 상기 GCC를 수행할 수 있다. 상기 XNOR는 비트(bit)간의 연산을 다루는 논리연산자 중에 하나로서, 서로의 비트가 같으면 1을 출력하고 다르면 0을 출력하는, 배타적 부정 논리합(exclusive NOR)를 의미한다.

Time delay,

_Ym is a digital acoustic signal, _xm is sampling data, i is a natural number of 2 or more, and N is the number of sampling data, to calculate a time delay corresponding to the digital acoustic signal . In particular, since the 1-bit digital sound signal is used here, the GCC can be performed using XNOR unlike the conventional case. The XNOR is one of logical operators for dealing with operations between bits, and means an exclusive NOR that outputs 1 if the bits are equal to each other, and outputs 0 when they are different from each other.

In the past, when 12-bit digital data is used, multiplication of 12-bit digital data is performed to apply the GCC. However, according to the acoustic processor 20 according to the embodiment of the present invention, The calculation can be performed simply and quickly.

Here, the sound processing unit 20 may be implemented as one physical device, for example, an FPGA (Field-Programmable Gate Array), or may be implemented using a plurality of physical devices. Accordingly, the sound processing unit 20 may be implemented using a separate physical device such as a sound signal input device and a signal processor.

The sound signal input device receives the analog sound signal and generates the 1-bit digital sound signal. The sound signal input device may be implemented by a dedicated IC (Integrated Circuit) or an operational amplifier. The signal processor may be implemented by a dedicated IC, a controller, etc., and the time delay of a sound wave input to each microphone 10 may be calculated by determining the matching of the digital sound signals.

Specifically, the acoustic signal input unit may sample the analog acoustic signal to generate sampling data. The acoustic signal input unit compares i-th sampled data with i-1 (i is a natural number of 2 or more) A 1-bit digital sound signal corresponding to the sound signal can be generated. For example, if the i-th sampling data is larger than the (i-1) -th sampling data, the digital sound signal may be generated by outputting 1 while otherwise outputting 0. Thereafter, the acoustic signal input device may transmit the digital acoustic signal to the signal processor.

The signal processor receiving the digital sound signal calculates a cross-correlation between the digital sound signals by using GCC (Generalized Cross-Correlation) for the digital sound signal, The time delay between the digital acoustic signals can be calculated using the cross-correlation value between the acoustic signals. Specifically, the signal processor may calculate the time delay using Equation (2).

Thereafter, it is possible to calculate the position of the sound source using the calculated time delay. The details of calculating the position of the sound source using the time delay will not be described in detail here.

In order to test the performance of the sound source location tracking apparatus according to an embodiment of the present invention, the experiment can be performed by providing a microphone and a sound source in a laboratory as shown in FIG. At this time, the microphones MIC1, MIC2, and MIC3 and the sound sources S1 to S18 may have a geometric structure as shown in FIG. 4A. The concrete positions of the microphones MIC1, MIC2, and MIC3 and the sound sources S1 to S18 may be represented by coordinates with respect to the origin as shown in FIG. 4B.

5 is a time delay measurement when the sound source exists at the position of S5. The time delay measured by the conventional sound source position tracking device and the time delay measured by the sound source position tracking device according to an embodiment of the present invention are shown as sample numbers (NS: Number of Samples). Here, the conventional sound source location tracking apparatuses calculate time delays using PC (Polarity Coincidence Algorithm), AMDF (Average Magnitude Difference Function), and GCC (time-domain generalized cross-correlation). Here, the sound source location tracking device according to an embodiment of the present invention corresponds to GCC-BC (Generalized Cross-Correlation-Binary Compared Acoustic Code). Referring to FIG. 5, although the apparatus for tracking a sound source according to an embodiment of the present invention reduces the conventional 12 bits to 1 bit, the calculated time lag is not substantially different from the conventional method .

6 is a comparison chart comparing the performance of a sound source position tracking apparatus according to an embodiment of the present invention with a conventional sound source position tracking apparatus according to a signal to noise ratio (SNR), and also shows a signal- It can be confirmed that the present invention (GCC-BC) is more advantageous.

Finally, FIG. 7 is a comparison table in which the time delay according to each sound source position is calculated, and it is confirmed that there is no difference in the calculated time delay value when compared with the case of using the conventional 12-bit .

8 is a flowchart illustrating a sound source position tracking method according to an embodiment of the present invention.

8, the sound source position tracking method according to an embodiment of the present invention includes a sound wave sensing step S110, a waveform extracting step S120, a time delay calculating step S130, and a position determining step S140 can do.

Hereinafter, a sound source position tracking method according to an embodiment of the present invention will be described with reference to FIG.

In the sound wave sensing step (S110), when a sound wave is applied, the sound wave sensing step S110 may generate an analog sound signal by sensing a sound wave applied to the plurality of microphones. As discussed above, the analog acoustic signal may be represented by an electrical signal corresponding to the vibration of the sound wave, and the analog acoustic signal may include a time delay depending on the distance between the sound source from which the sound wave is generated and the microphone have.

In the waveform extraction step (S120), the sound processing unit may generate a digital sound signal corresponding to the waveform of the analog sound signal. When the time delay between the analog acoustic signals is grasped, the position of the sound source generating the sound waves can be determined. In the waveform extraction step (S120), the analog sound signal may be firstly converted into a digital sound signal in order to easily calculate the time delay between the analog sound signals. Specifically, the waveform extracting step (S120) may generate sampling data by sampling the analog acoustic signals. In particular, the waveform extracting step (S120) may generate the digital acoustic signals by comparing sizes between adjacent sampling data have. In particular, if the i-th input sampling data is larger than the (i-1) -th input sampling data, the digital sound signal may be generated by outputting 1 and outputting 0 otherwise. At this time, i is a natural number of 2 or more. That is, whether the analog acoustic signal is increased or decreased can be represented by the digital acoustic signal, thereby obtaining information on the external appearance of the analog acoustic signal. In this case, the digital acoustic signal may be represented by a 1-bit binary sequence expressed by 0 or 1.

Specifically,

y _m is a digital sound signal, x _m is sampling data, and i is a natural number of 2 or more. x ₁ denotes sampling data for an analog sound signal input from the first microphone, and x ₂ denotes sampling data for an analog sound signal input from the second microphone.

The time delay calculation step (S130) may calculate the time delay of the sound wave applied to the microphone using the digital sound signal. In order to calculate the time delay, GCC may be applied to the 1-bit digital sound signal,

Specifically,

는 시간지연,

는 디지털음향신호의 상호상관값, y_m은 디지털음향신호, x_m은 샘플링 데이터, i는 2 이상의 자연수, N은 샘플링 데이터의 개수를 이용하여 상기 디지털음향신호에 대응하는 시간지연을 계산할 수 있다. 특히, 여기서는 1bit의 디지털음향신호를 활용하므로, 종래와 달리 XNOR를 활용하여 상기 GCC를 수행할 수 있다. 상기 XNOR는 비트(bit)간의 연산을 다루는 논리연산자 중에 하나로서, 서로의 비트가 같으면 1을 출력하고 다르면 0을 출력하는, 배타적 부정 논리합(exclusive NOR)를 의미한다.

Time delay,

_Ym is a digital acoustic signal, _xm is sampling data, i is a natural number of 2 or more, and N is the number of sampling data, to calculate a time delay corresponding to the digital acoustic signal . In particular, since the 1-bit digital sound signal is used here, the GCC can be performed using XNOR unlike the conventional case. The XNOR is one of logical operators for dealing with operations between bits, and means an exclusive NOR that outputs 1 if the bits are equal to each other, and outputs 0 when they are different from each other.

The location determination step 140 may determine the location of the sound source using the time delay. As described above, since the information on the position of the sound source can be obtained by calculating the time delay, it is possible to determine the position of the sound source using the time delay. The details of calculating the position of the sound source using the time delay will not be described in detail.

The present invention is not limited to the above-described embodiments and the accompanying drawings. It will be apparent to those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

10: microphone 20: sound processor
S: Sound source
S110: Sound wave detection step S120: Waveform extraction step
S130: Time delay calculation step S140: Position determination step

Claims (8)

A plurality of microphones for sensing an applied sound wave and generating analog sound signals, respectively; And
And a sound processing unit for generating a digital sound signal corresponding to a change in size of a waveform of the analog sound signal and calculating a time delay of a sound wave applied to each of the microphones using the digital sound signal,
Wherein the acoustic processing unit generates the digital acoustic signal by sampling the analog acoustic signal to generate sampling data and comparing sizes of adjacent sampling data.
delete The sound processing apparatus according to claim 1, wherein the sound processing unit
1 if the i-th sampled data is larger than the (i-1) th sampled data, and outputs 0 if the i-th sampled data is less than or equal to the (i-1) And i is a natural number of 2 or more.
The sound processing apparatus according to claim 1, wherein the sound processing unit

y _m is a digital acoustic signal, x _m is sampling data, and i is a natural number of 2 or more to generate the digital acoustic signal of 1 bit consisting of 0 and 1.
The sound processing apparatus according to claim 1, wherein the sound processing unit
Calculating a cross-correlation between each of the digital acoustic signals using GCC (Generalized Cross-Correlation), and calculating a time delay corresponding to each of the digital acoustic signals using the cross-correlation value A sound source position tracking device.
6. The sound processing apparatus according to claim 5,

Time delay,

_Ym is a digital acoustic signal, _xm is sampling data, i is a natural number of 2 or more, N is the number of sampling data, and calculates the time delay corresponding to the digital acoustic signal Sound source location tracking device.
The sound processing apparatus according to claim 1, wherein the sound processing unit
And outputs 1 if the i-th input sampling data is greater than i-1 (i is a natural number equal to or greater than 2) input sampling data, and the i-th input sampling data is i An acoustic signal input unit for outputting 0 when the sampling data is smaller than or equal to -1th sampling data to generate a 1-bit digital acoustic signal corresponding to each of the analog acoustic signals; And
Calculating a cross-correlation between the digital acoustic signals by applying GCC (Generalized Cross-Correlation) to each of the digital acoustic signals, and using the cross-correlation values to correspond to the respective digital acoustic signals And a signal processor for calculating a time delay of the sound source.
A sound wave sensing step of sensing, when a sound wave is applied, sound waves to which a plurality of microphones are applied, respectively, to generate an analog sound signal;
A waveform extracting step of the acoustic processing unit generating a digital acoustic signal corresponding to the waveform of the analog acoustic signal;
Calculating a time delay of a sound wave applied to the microphone using the digital sound signal; And
And a position determining step of determining a position of the sound source using the time delay,
Wherein the waveform extracting step includes sampling the analog acoustic signal to generate sampling data, and comparing sizes of adjacent sampling data to generate the digital acoustic signal.

Images