close

Вход

Забыли?

вход по аккаунту

?

JP2011122854

код для вставкиСкачать
Patent Translate
Powered by EPO and Google
Notice
This translation is machine-generated. It cannot be guaranteed that it is intelligible, accurate,
complete, reliable or fit for specific purposes. Critical decisions, such as commercially relevant or
financial decisions, should not be based on machine-translation output.
DESCRIPTION JP2011122854
The present invention displays a direction of arrival of a correctly estimated sound and a
direction of arrival of a sound which can not be correctly estimated in a distinguishable manner.
SOLUTION: A sound arrival direction determination system 100 estimates a sound arrival
direction by a sound source position estimation unit 24 of an arithmetic processing unit 23 of a
personal computer 20 from a difference in arrival time of sound to each microphone of a
measurement unit 10. The virtual sound source position determination unit 25 determines the
authenticity of the estimated direction of arrival of the sound. The display processing unit 26
simultaneously displays the sound source position determined to be low in authenticity and the
sound source position determined to be high in authenticity on the display 29 in a
distinguishable manner. [Selected figure] Figure 1
Sound arrival direction determination system and program
[0001]
The present invention relates to a system for estimating and displaying the position of a sound
source, and more particularly to a sound arrival direction determination system and program
capable of determining the credibility of the estimated sound arrival direction and determining
the sound arrival direction.
[0002]
Conventionally, as a system for estimating the position of a sound source such as noise, for
example, a sound source tracking system disclosed in Patent Document 1 is known.
04-05-2019
1
In this sound source tracking system, for example, microphones are respectively disposed at five
apexes of a quadrangular pyramid, and from the output signals of these five microphones, the
arrival time difference of sound of the target frequency to the five microphones is determined ,
Estimate the direction of arrival of sound for each frequency. Then, in the image captured by an
imaging means such as a CCD camera, the estimated direction of arrival of the sound is displayed
as a sound source position, for example, changed in size according to the sound pressure level,
and the noise source can be visually grasped I am trying to
[0003]
However, the above method assumes that there is only one sound source position, and when
there are multiple sound sources of the same frequency or when the influence of the reflected
sound is large, it is necessary to correctly estimate the arrival direction of the sound. It will be
difficult. Therefore, a method of separating multiple sound sources as disclosed in, for example,
Patent Document 2 is also known. In this separation method, for example, when the positions of
a plurality of sound sources having the same or close frequencies are separated and identified,
the amplitude of the sound from each sound source with respect to the amplitude of the
observation sound input to the microphone serving as the reference. A complex vector is
assumed for each sound source which represents the magnitude of and whose angle represents
the phase difference with respect to the observation sound. Then, after estimating the direction
of each sound source using the set of assumed complex vectors, the phase difference of the
sound reaching each microphone from a specific sound source in the estimated sound source is
determined, and a new complex is obtained from this phase difference. A vector is determined,
and a complex vector is calculated such that the vector of the difference between the new
complex vector and the complex vector of the specific sound source assumed is the smallest. The
directions of the sound sources are separated and specified as the estimation directions of the
respective sound sources, in which the direction of each sound source providing the set of
complex vectors including the complex vector thus obtained is determined.
[0004]
Furthermore, for example, a sound source position estimation system disclosed in Patent
Document 3 is also known. In this sound source position estimation system, when the specific
microphones are the microphones M5, the phase difference between the sound pressure signal
observed by each of the microphones M1 to M4 and the sound pressure signal of the specific
04-05-2019
2
microphone M5 and the sound transmission from the estimated sound source direction And the
phase difference between the sound pressure signal input to each of the microphones M1 to M4
and the sound pressure signal of the specific microphone M5, assuming that it is calculated, and
the difference between the phase differences exceeds a predetermined angle. Determines that the
estimated sound source direction is not appropriate, and discards the data of the estimated sound
source direction.
[0005]
JP 2003-111183 JP JP 2007-96418 JP JP 2008-224259 JP
[0006]
However, in the sound source tracking system disclosed in Patent Document 1, as described
above, when there are a plurality of sound sources of the same frequency, or when the influence
of the reflected sound due to the road surface or wall is large, The direction of arrival of can not
be estimated.
[0007]
Further, according to the separation method disclosed in Patent Document 2 above, although it is
possible to separate synthetic sounds of the same frequency if only two sounds are synthesized,
in an actual on-site environment, such simple Synthetic sound is not a source of noise, and it
must be said that its practicality is poor.
[0008]
Furthermore, in the sound source position estimation system disclosed in Patent Document 3
above, data of the estimated sound source direction when it is determined that the sound source
direction is not appropriate is discarded, so the search operation of the actual sound source
position is efficiently performed. I can not
[0009]
An object of the present invention is to provide a sound arrival direction determination system
and program capable of solving the problems due to the prior art described above and
performing the search operation of the actual sound source position accurately and efficiently.
[0010]
In order to solve the above problems and achieve the object, according to the sound arrival
04-05-2019
3
direction determination system of the present invention, a plurality of sound collection means
arranged at a predetermined interval and a difference in arrival time of sound between the
plurality of sound collection means Arrival direction estimation means for estimating the arrival
direction of sound, authenticity determination means for determining the authenticity of the
arrival direction of sound estimated by the arrival direction estimation means, and arrival of
sound to be estimated by the arrival direction estimation means Information indicating the
arrival direction of the sound estimated by the arrival direction estimation means together with
the authenticity determined by the authenticity determination means, on the image captured by
the imaging means And display means for displaying.
[0011]
In the sound arrival direction determination system according to the present invention, the
plurality of sound collection means may preferably be at least eight microphones arranged at the
apex of a cube.
[0012]
Also, the arrival direction estimation means is based on the phase spectrum of the frequency of
the sound whose estimation direction is observed by each of the microphones, from the time
difference of arrival time of the sound between the four microphones arranged in the horizontal
plane. The direction of arrival of the sound can be estimated, and the direction of arrival of the
sound in the elevation direction can be estimated from the difference in time of arrival of the
sound between the four microphones arranged in the vertical plane.
[0013]
Further, the arrival direction estimation means is configured such that a plane formed by four
microphones of a cube formed by eight microphones forms an angle of 45 ° with the estimated
direction of arrival of sound in the horizontal direction. The smaller surface, and the closest
surface as viewed from the horizontal sound arrival direction, can be used to calculate the
elevation angle.
[0014]
Preferably, the credibility determining means determines a virtual sound source position from
the estimated direction of arrival of the sound, and phase differences and amplitude ratios among
the plurality of sound collecting means determined from the virtual sound source position; The
phase difference and the amplitude ratio among the plurality of sound collecting means are
compared with each other based on the measured value measured by the sound collecting
means, and the credibility of the direction of arrival of the sound is determined based on the
04-05-2019
4
comparison result. be able to.
[0015]
Further, the display means may display, on the image, information indicating the arrival direction
of the sound determined to be high in credibility by the credibility determination means and the
arrival direction of the sound determined to be low in credibility. It can be configured to be
displayed simultaneously in a distinguishable state.
[0016]
Further, the sound arrival direction determination program according to the present invention
captures outputs from a plurality of sound collection means arranged at predetermined intervals,
and captures images of the sound arrival direction to be estimated from the image pickup means,
It is a sound arrival direction determination program that causes a computer to execute a process
of estimating the arrival direction of the sound from the output of the sound collection means
and a process of determining the authenticity of the estimated direction of the sound. A process
of estimating the direction of arrival of the sound from the time of arrival of the sound between
the plurality of sound collection means, a process of determining the authenticity of the direction
of arrival of the estimated sound, and the process of determining the estimated direction of
arrival of the sound And causing the computer to execute a process of displaying the information
shown together with the credibility on the image captured from the imaging means.
[0017]
According to the present invention, since information indicating the estimated direction of arrival
of the sound and the authenticity thereof is displayed on the image captured by the imaging
means, it can be understood that an imaginary sound source exists on the image. .
Therefore, the search operation of the actual sound source position is carried out accurately and
efficiently by continuing the measurement by moving the sound collection means or changing
the direction until the virtual sound source is determined as the actual sound source. be able to
[0018]
It is a whole block diagram of the arrival direction determination system of the sound which
04-05-2019
5
concerns on one Embodiment of this invention.
It is a perspective view which shows the example of arrangement | positioning of the sound
collection means of the determination system.
It is a functional block diagram of the part which performs arithmetic processing of the
determination system.
It is a figure which shows an example of the display screen in the determination system.
It is a figure for demonstrating the procedure which estimates the arrival direction of the
elevation angle direction of the sound in the determination system.
[0019]
Hereinafter, with reference to the accompanying drawings, a sound arrival direction
determination system (hereinafter simply referred to as a “determination system”) according
to an embodiment of the present invention.
Will be described in detail.
FIG. 1 is a diagram showing an entire configuration of a determination system.
[0020]
A determination system 100 according to the present embodiment includes a measurement unit
10 for collecting and imaging a sound source search space, a microphone amplifier 11 for
processing an audio signal and an image signal acquired by the measurement unit 10, and a low
pass filter (LPF 12, interface means such as an A / D converter 13 and a video capture device 14,
and a personal computer 20 for taking in and processing audio information and video
information through the interface means.
04-05-2019
6
[0021]
The measurement unit 10 picks up a plurality of omnidirectional microphones M1 to M8 as
sound collection means arranged at predetermined intervals, a microphone frame 15 for
supporting the microphones M1 to M8, and an image in front of a lens It comprises a camera 19
and a base 16 for supporting a rotating frame 18 mounted with the microphones M1 to M8 and
the camera 19 together with a support member 17 consisting of a tripod.
When the measurement unit 10 is configured as described above, measurement is performed by
moving the measurement position of the measurement unit 10 at a plurality of locations, or
measurement is performed at a plurality of angles by rotating the rotating frame 18 at the same
measurement location. can do.
Although not shown, the measurement unit 10 may further include a GPS device for measuring
the absolute position of each of the microphones M1 to M8 on the ground.
[0022]
As also shown in FIG. 2, in the measurement unit 10, eight microphones M1 to M8 are installed
upward at the apex of a cube whose one side is a distance L.
The distance L needs to be at least smaller than the half wavelength of the sound to be searched,
and is set to, for example, several centimeters to several tens centimeters.
Then, when the imaging direction of the camera 19 is forward, a direction perpendicular to the
side surface formed by the microphones M1, M2, M6, and M7 of the cube is the X direction, and
the front surface formed by the microphones M1, M4, M8, and M5. Direction is defined as Y
direction, and a direction perpendicular to the upper surface formed by microphones M1, M2,
M3 and M4 is defined as Z direction, and the imaging direction of camera 19 is defined as Y
direction and horizontal angle (both “horizontal angle” It is called.
) Θ is an inclination from the YZ plane to the X direction, and an elevation angle (also referred to
as “elevation direction angle”).
04-05-2019
7
) Is the inclination from the XY plane to the Z direction. Therefore, the Y direction is the
horizontal angle θ = 0 ° and the elevation angle φ = 0 °, the X direction is the horizontal
angle θ = 90 °, the elevation angle φ = 0 °, and the Z direction is the horizontal angle θ = 0
° and the elevation angle φ = 90 ° It becomes.
[0023]
The acoustic signals output from the microphones M1 to M8 are amplified by the microphone
amplifier 11, and the high frequency is suppressed by the LPF 12. The LPF 12 functions as an
anti-aliasing filter that prevents aliasing error (Aliasing) when converting an acoustic signal into a
digital signal. The output of the LPF 12 is converted into a digital signal by the A / D converter
13 and input to the personal computer 20. On the other hand, the video signal from the camera
19 is once stored in the video capture device 14 and then taken into the personal computer 20.
The personal computer 20 is provided with a display 29 for displaying video information.
[0024]
The personal computer 20 includes, as shown in FIG. 3, a storage unit 22 for storing various
information including an input unit 21 such as a keyboard, a mouse, and a joystick, parameters
received by the input unit 21, and the like. The arithmetic processing unit 23 performs various
arithmetic processing based on information from the input unit 21 and acoustic information and
image information from the measurement unit 10.
[0025]
Note that the arithmetic processing unit 23 of the personal computer 20 is a hardware such as a
CPU, a RAM, a ROM, etc., and executes software such as a sound arrival direction determination
program by using a sound source position estimation unit 24 and a virtual sound source position
determination unit 25. And functions as the display processing unit 26.
[0026]
The sound source position estimation unit 24 estimates the sound source position from the
arrival time difference of the sound between the microphones M1 to M8 based on the acoustic
information from the measurement unit 10 to estimate the arrival direction of the sound.
04-05-2019
8
The virtual sound source position determination unit 25 determines the authenticity of the
estimation result estimated by the sound source position estimation unit 24 to determine the
virtual sound source position.
The display processing unit 26 receives, from the measurement unit 10, information indicating
the arrival direction of the sound estimated by the sound source position estimating unit 24 and
the credibility thereof based on the determination result determined by the virtual sound source
position determining unit 25. A process is performed to display on the display 29 together with
the video information.
[0027]
Regarding the microphones M1 to M8 of the measurement unit 10, the phase difference
characteristics with respect to one microphone are measured in advance, and eight microphones
having small measured phase difference characteristics are appropriately combined to arrive the
sound. Suppose that it is configured to determine the direction. For example, when the
microphone M1 is one microphone, for the microphones M2 to M8, phase difference information
and amplitude ratio information of all frequencies to be measured for the microphone M1 are
measured, and the input unit 21 is used. These pieces of information are stored in the storage
unit 22 of the personal computer 20 via the above.
[0028]
Next, the process of determining the direction of arrival of sound using the determination system
100 configured as described above will be described. First, an acoustic signal which is an analog
signal collected by the microphones M 1 to M 8 of the measurement unit 10 and amplified by the
microphone amplifier 11 is filtered by the LPF 12.
[0029]
The filtered acoustic signal is converted into a digital signal sampled (sampled) at a
predetermined cycle by the A / D converter 13 and stored as acoustic information in the storage
04-05-2019
9
unit 22 of the personal computer 20. The sound source position estimation unit 24 in the
arithmetic processing unit 23 of the personal computer 20 reads the acoustic information of
each of the microphones M1 to M8 stored in the storage unit 22 and generates a frame of a
predetermined length.
[0030]
Specifically, the sound source position estimation unit 24 frames the read out acoustic
information in units of a predetermined length of 0.5 sec to 1.0 sec, for example. Note that the
sound source position estimation unit 24 performs window function processing such as a
Hamming window or a Hanning window, for example, on each of the generated frames, and
performs an FFT (fast Fourier transform) on an audio signal of each microphone M1 to M8. The
conversion is processed into a complex exponential function, and the result is stored in the
storage unit 22.
[0031]
Then, for example, the complex conjugate elements of the result of the FFT processing of the
sound information of the frame unit of one microphone M1 and the result of the FFT processing
of the sound information of each of the other microphones M2 to M8 are multiplied by each
other , The phase spectrum of the multiplication result of the elements of this complex conjugate
is obtained.
[0032]
Furthermore, the phase difference information of each of the microphones M2 to M8 with
respect to the microphone M1 stored in the storage unit 22 is read out, correction processing of
the phase spectrum is performed, and the phase difference (rad) of the microphones M2 to M8 is
obtained.
Next, the following processing is performed on the frequency f (Hz) for which it is desired to
estimate the arrival direction of sound. However, the frequency f must be equal to or less than
the frequency at which the distance L between the microphones M1 to M8 is a half wavelength,
and it is assumed that the sound wave is a plane wave.
04-05-2019
10
[0033]
First, the sound source position estimation unit 24 divides, for example, the phase difference
between frame units of the microphone M1 and the microphone M3 by 2πf to obtain the arrival
time difference D13 (sec) between the microphone M1 and the microphone M3. Next, the phase
difference of the microphone M1 and the microphone M2 is subtracted from the phase
difference of the microphone M1 and the microphone M4 in frame units, and the value is divided
by 2πf to obtain the arrival time difference D24 between the microphone M2 and the
microphone M4. .
[0034]
After the arrival time difference with respect to these microphones M1 to M4 is obtained, the
horizontal direction angle θ (°) of the sound of frequency f is estimated from the following
equation (1).
[0035]
[0036]
Based on the horizontal angle θ of the sound of frequency f thus estimated, the combination of
microphones M1 to M8 for estimating the elevation direction of the sound is determined, and the
arrival time difference between the microphones is determined in the same manner as described
above. The elevation angle angle φ (°) of the frequency f is estimated by (2) to (5).
[0037]
[0038]
[0039]
[0040]
[0041]
04-05-2019
11
In the above equations (2) to (5), Dij (i, j = 1 to 8) represents the arrival time difference between
the microphone Mi and the microphone Mj.
[0042]
Formulas (2) to (5) shown in the above I to IV, as shown in FIG. 5, are surfaces formed by four
microphones used for calculation of the elevation angle .phi. Among cubes formed by the
microphones M1 to M8. Is an angle smaller than 45 ° with respect to the estimated horizontal
sound arrival direction (horizontal angle θ), and viewed from the estimated horizontal sound
arrival direction (horizontal angle θ) It is characterized in that it is the closest surface.
As described above, by calculating the elevation angle φ with the four microphones constituting
the surface satisfying the relationship as described above according to the horizontal direction
angle θ, the estimation accuracy is far more remarkable than the conventional example using
the five microphones. improves.
[0043]
That is, the present inventors conducted an experiment to verify the validity of the sound source
direction estimated by a system using five conventional microphones.
As a result, a sound source which is obviously not a real sound source (hereinafter referred to as
"imaginary sound source".
Position (hereinafter referred to as "imaginary sound source position").
It has been recognized that in some cases, it may be misjudged as an actual sound source
position with extremely high relevance (high reliability).
It has been found that such an erroneous determination is particularly noticeable when there are
reflecting surfaces in the vertical direction.
04-05-2019
12
In this regard, it has been confirmed that the use of this system reduces such false positives.
[0044]
After estimating the horizontal angle θ and elevation angle φ which are directions of arrival of
the sound of frequency f by the sound source position estimation unit 24, next, the authenticity
of the direction of arrival of the sound of frequency f estimated by the virtual sound source
position determination unit 24 is determined .
That is, the virtual sound source position determination unit 25 reads the result of the FFT
processing of the frame unit sound signal of each of the microphones M1 to M8 obtained by the
sound source position estimation unit 24 from the storage unit 22 and obtains each amplitude
spectrum.
[0045]
Then, the amplitude spectrum of the acoustic signal in frame units of the microphones M2 to M8
is corrected using the amplitude ratio information of each of the microphones M2 to M8 with
respect to the microphone M1 stored in the storage unit 22, and then the amplitude of the
microphone M1 is obtained. Divide by the spectrum.
Thereby, the amplitude ratio of the microphones M2 to M8 is determined.
[0046]
Next, the virtual sound source position determination unit 25 uses appropriate distance (for
example, 100 m) information given by an operation input to the input unit 21 or the like as the
distance to the sound source in the arrival direction of the estimated sound. The virtual sound
source position is determined from the three-dimensional orthogonal coordinate values (X, Y, Z)
by the following equations (6) to (8).
[0047]
04-05-2019
13
[0048]
[0049]
[0050]
Here, rn indicates the distance to the sound source given in advance.
Then, based on the three-dimensional orthogonal coordinate value of the virtual sound source
position and the three-dimensional orthogonal coordinate value of each of the microphones M1
to M8, the distance (m) to each microphone M1 to M8 is calculated from the calculated virtual
sound source position The distance difference (m) of the microphones M2 to M8 with respect to
the microphone M1 is determined from the distance.
The distance r3 between the three-dimensional orthogonal coordinates (X1, Y1, Z1) and (X2, Y2,
Z2) can be obtained by the following equation (9).
[0051]
[0052]
The value obtained by dividing the distance difference thus determined by the sound velocity (m
/ sec) is multiplied by -2πf, and the phase difference of the microphones M2 to M8 with respect
to the microphone M1 is calculated as a theoretical value.
Such a method of obtaining the phase difference by a theoretical value is a method that is useful
even when the arrangement mode or the number of microphones changes.
[0053]
Then, the virtual sound source position determination unit 25 sets the phase difference of the
04-05-2019
14
microphones M2 to M8 with respect to the microphone M1 theoretically determined from the
arrival direction of the sound estimated by the sound source position estimation unit 24 to be a
first angle of 1 A phase difference between microphones M2 to M8 with respect to microphone
M1 determined by multiplication of complex vectors and elements of complex conjugates
obtained as a result of FFT processing from actual measured values as described above is an
angle and an amplitude ratio to microphone M1 is a length Define two complex vectors.
[0054]
Then, the distance between the end point of the first complex vector defined from the estimated
sound arrival direction and the end point of the second complex vector defined by actual
measured values (measured values) for the microphones M2 to M8, respectively, Each of the
microphones M2 to M8 is obtained and summed, and then divided by the frequency f.
The distance r2 between the end points of two complex vectors can be obtained by the following
equation (10).
[0055]
[0056]
Where α1 is the length of the first complex vector, ψ1 is the angle of the first complex vector,
α2 is the length of the second complex vector, and ψ2 is the angle of the second complex
vector, Each is shown.
A value normalized by a frequency of 1000 Hz can be obtained by multiplying, for example,
1000 by a value obtained by division using the frequency f as described above, and this value
can be a numerical value for determining a virtual sound source (determination Value).
Note that the value divided by the frequency f may be determined for each of the microphones
M2 to M8, and the determination value may be determined by the averaged value.
04-05-2019
15
[0057]
Here, this determination value is referred to as a microphone complex vector deviation value, and
it can be determined that the reliability is higher as the value is closer to 0.
Specifically, according to the experiment of the applicant, it has been found that if the
microphone complex vector deviation value is about 0.1 or less per microphone, it can be judged
that the credibility is high.
[0058]
As described above, the determination system 100 according to the present embodiment can
perform the virtual sound source determination by using the distance between the complex
vectors even when there is a phase difference between the microphones and there is an
amplitude difference between the microphones. System. Further, in this determination system
100, by dividing the distance between complex vectors by the frequency, it is possible to use the
microphone complex vector divergence value as a common value at all frequencies for evaluation
of authenticity.
[0059]
The display processing unit 26 uses the image information captured by the camera 19 on the
display 29 based on the estimated direction of arrival of the sound at the frequency f thus
obtained, the amplitude spectrum, and the microphone complex vector deviation value.
Information indicating the sound source position of the frequency f is drawn, for example,
graphically at the corresponding position on the image to be displayed on the screen.
[0060]
Specifically, a sound source position determined by the virtual sound source position
determination unit 25 to have a large microphone complex vector deviation value and low
credibility is displayed as, for example, a rectangular figure so that the rectangular inner region
is transparent. draw.
04-05-2019
16
Further, the sound source position determined to have a small microphone complex vector
deviation value and high credibility is displayed as, for example, a circular figure, and the circular
inner area is subjected to a filling process or a shading process. Draw for easy visual recognition.
[0061]
Furthermore, if the size of the figure drawn in proportion to the amplitude spectrum amount is
enlarged or the display color of the figure is changed according to the frequency f for drawing,
the actual sound source position and the virtual sound source position can be easily identified. It
is possible to In addition, for example, the microphone complex vector deviation value is
provided with fine steps, and drawing is performed with a figure having a large number of angles
so as to be triangle → square → pentagon ... in order from the lowest reliability to the most
reliable. The high ones may be drawn in a circle or the like.
[0062]
In the example of the display screen of the display 29 shown in FIG. 4, for example, the display
color is changed according to the frequency f, the size of the display area is changed according to
the amplitude spectrum amount, and the sound source position determined to be less credibility
is shown as a rectangular figure. And a drawing example in the display processing unit 26 in
which the sound source position determined to be high in credibility is represented by a circular
figure. According to this example, the sound sources B to D are displayed as rectangular figures,
and the sound source A is displayed as a circular figure larger than these. Therefore, it can be
seen at a glance that the amplitude spectrum amount of the sound source A is larger than the
sound sources B to D, the display position of the sound source A is the actual sound source
position, and the display position of the sound sources B to D is the imaginary sound source
position .
[0063]
Note that the horizontal direction angle θ and the elevation angle direction angle φ of the
direction of arrival of the sound estimated as described above are directions from the position of
the center of gravity of the cube formed by the microphones M1 to M8, so the approximate
distance rn to the sound source If is known, it can also be drawn as follows. That is, the display
processing unit 26 sets the distance difference in the front-rear direction of the image center to
04-05-2019
17
the center of gravity position of the cube constituted by the microphones M1 to M8 as PY, the
distance difference in the left-right direction as PX, and the distance difference in the vertical
direction as PZ. It is also possible to convert and draw in the horizontal direction PH and
elevation direction PE from the image center indicated by the image information captured by the
camera 19 using the following equations (11) and (12).
[0064]
[0065]
[0066]
As described above, according to the determination system 100 according to the present
embodiment, the direction of arrival of sound is estimated to determine the credibility thereof,
the virtual sound source determination is performed, and the actual sound source position and
the virtual sound source position are described above. Can be simultaneously displayed on the
captured image.
As a result, even if there are multiple sound sources with the same frequency or the influence of
the reflected sound from the road surface or wall is large and the arrival direction of the sound of
frequency f can not be estimated correctly, the sound of frequency f is Although it exists, it can
be visually recognized that the estimated direction of arrival of the sound is low in authenticity.
In this case, if the measurement position of the measurement unit 10 is moved and changed until
the reliability of the estimated direction of arrival of the sound becomes high, it becomes possible
to search for the actual sound source position.
[0067]
The present invention is not limited to the embodiments described above. For example, in the
above embodiment, the directions of arrival of sound are estimated using eight microphones M1
to M8 arranged at each vertex of a cube, but a plurality of sets of such eight microphones M1 to
M8 (for example, four) The determination accuracy is further improved by arranging the sets at
04-05-2019
18
appropriate intervals and performing the verification process using the determination results
obtained for each set. Further, at this time, by making the intervals of the microphones M1 to M8
of each set different, it is possible to make the frequency bands that can be grasped different for
each group. That is, the set of microphones M1 to M8 with a wide interval can estimate the
arrival direction of lower frequency sound.
[0068]
The sound arrival direction determination system according to the present invention is
particularly useful for a sound source search system for searching for a true sound source
position, such as identification of an abnormal noise occurrence point of a car.
[0069]
DESCRIPTION OF SYMBOLS 10 measurement unit 11 microphone amplifier 12 low pass filter 13
A / D converter 14 video capture device 15 microphone frame 16 base 17 support member 18
rotation frame 19 camera 20 personal computer 21 input unit 22 storage unit 23 arithmetic
processing unit 24 sound source position estimation Part 25 Virtual sound source position
judgment part 26 Display processing part 29 Display 100 Sound arrival direction judgment
system
04-05-2019
19
Документ
Категория
Без категории
Просмотров
0
Размер файла
31 Кб
Теги
jp2011122854
1/--страниц
Пожаловаться на содержимое документа