close

Вход

Забыли?

вход по аккаунту

?

JP2016127430

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 JP2016127430
Abstract: To provide a sound source direction estimation device capable of accurately estimating
the direction of a sound source without lowering the upper limit of the measurable frequency.
SOLUTION: First to fourth microphones M1 to M4 which are not in line with each other are
arranged on a plane plate 12, and a diagonal of a quadrangle which has the positions of the first
to fourth microphones M1 to M4 as apexes. A fifth microphone M5 is disposed at the
intersection, and among each of the two microphone pairs which are different from one another
and which are different from each other and which are different from each other and which are
selected from these five microphones M1 to M5, From the arrival time difference of the sound
pressure signal input to each of the microphones constituting one microphone pair, the arrival
time difference of the sound pressure signal input to each of the microphones constituting the
other microphone pair, the position coordinates of the microphones and the velocity of sound It
was made to estimate the direction of the sound source. [Selected figure] Figure 1
Sound source direction estimation device
[0001]
The present invention relates to an apparatus for estimating the direction of a sound source from
information of sound collected by a plurality of microphones, and more particularly to a sound
source direction estimation apparatus having a configuration in which a plurality of microphones
are arranged on a plane.
[0002]
Conventionally, as shown in FIG. 5, four microphones M51 to M54 are arranged at
predetermined intervals on two straight lines orthogonal to each other, and a microphone M55
03-05-2019
1
for elevation angle estimation is a square formed by the microphones M51 to M54 as a bottom
surface. The sound pressure means of the sound propagating from the sound source is detected
by the sound collecting means 50 arranged at the position of the apex of the square pyramid to
be reached, and the arrival equivalent to the phase difference between the two microphones (Mi,
Mj) which make a pair There is known a method of estimating a horizontal angle θ and an
elevation angle φ which are directions of a sound source from a time difference D ij (see, for
example, Patent Document 1).
[0003]
JP, 2011-238985, A
[0004]
In the above-mentioned conventional method, the sound source direction and the size of the
arriving sound can be measured for each frequency, so information of the sound source can be
grasped with certainty, but in narrow spaces such as indoors, the influence of the reflected sound
from the wall etc. Because of this, it is necessary to use arithmetic processing to distinguish
between direct sound and reflected sound.
Therefore, if multiple microphones are installed on a flat plate and the sound pressure signal
input to the microphones is limited to only 180 ° in the front, which is the shooting direction of
the camera, the effect of the reflected sound from the back or side wall etc. Can be significantly
reduced, so that the estimation accuracy of the sound source direction can be greatly improved.
[0005]
By the way, in the estimation of the sound source direction, it is expected that the estimation
accuracy of the sound source direction is improved because the error in measurement is
relatively reduced if the microphone interval is increased.
Therefore, when estimating the direction of the sound source emitting a low frequency, the
microphone interval is expanded.
03-05-2019
2
However, simply increasing the distance between the microphones lowers the upper limit of the
frequency that can be measured, which makes it difficult to accurately measure the direction of a
sound source having a high frequency.
[0006]
The present invention has been made in view of the conventional problems, and it is an object of
the present invention to provide a sound source direction estimation device capable of accurately
estimating the direction of a sound source without lowering the upper limit of the measurable
frequency.
[0007]
According to the present invention, there is provided a sound source direction comprising sound
collection means for collecting a sound pressure signal of sound propagated from a sound
source, and sound source direction estimation means for estimating the direction of the sound
source from the sound pressure signal collected by the sound collection means. The estimation
apparatus, wherein the sound collecting means collects only the sound pressure signal of the
sound transmitted from the front, which is one side of the plane including the plane plate and the
plane plate installed on the plane plate. A fifth microphone arranged at the intersection of
diagonals of first to fourth microphones not exactly in line with each other and a quadrangle
(strictly speaking, a convex quadrilateral) whose apexes are positions of the first to fourth
microphones. And the sound source direction estimation means selects one of the microphones
of the two microphone pairs for each of the two microphone pairs which are selected from the
five microphones and are not in line with each other. From the arrival time difference of the
sound pressure signal input to each microphone forming the phone pair, the arrival time
difference of the sound pressure signal input to each microphone forming the other microphone
pair, the position coordinates of the microphone, and the sound velocity It is characterized by
estimating the direction of.
This makes it possible to construct three microphone pairs which are not in line with one another
and which have different microphone intervals L S, L M and L L (L S <L M <L L) from five
microphones M1 to M5. . Therefore, for a sound source direction that generates low frequency
sound, estimation is performed using data of two microphone pairs with a long microphone
distance (L = L L), and for a sound source direction that generates high frequency sound, the
microphone If estimation is performed using data of two microphone pairs with a short interval
(L = L s), the direction of the sound source can be accurately estimated without lowering the
upper limit of the measurable frequency.
03-05-2019
3
[0008]
Further, according to the present invention, the square is a square in which two diagonal lines
are orthogonal to each other, and a microphone pair constituting one diagonal line is a first
microphone pair, and a microphone pair constituting the other diagonal line is a fourth one.
When two microphone pairs are used, the second microphone pair is disposed at a position
where the first microphone pair is rotated by + 90 ° or -90 ° around the fifth microphone. I
assume. As a result, all of the three microphone pairs having different microphone spacings L S, L
M and L L (L S <L M <L L) become microphone pairs arranged on straight lines orthogonal to
each other. The estimation calculation of the direction can be easily performed. In addition to the
three microphone pairs having different microphone spacings L S, L M and L L (L S <L M <L L)
because the square is a square, two microphone pairs in the same axial direction are You can get
it. Therefore, the averaging accuracy of the sound source direction estimated by these
microphone pairs can further improve the estimation accuracy of the sound source direction.
[0009]
Further, the present invention comprises sound collection means for collecting a sound pressure
signal of sound propagated from a sound source, and sound source direction estimation means
for estimating the direction of the sound source from the sound pressure signal collected by the
sound collection means. A sound source direction estimation device, wherein the sound collecting
means is provided on a flat plate and the flat plate and picks only a sound pressure signal of
sound transmitted from the front which is one side of a plane including the flat plate. , The first
to fourth microphones disposed at each vertex of the square, wherein the sound collecting means
is selected from the four microphones, for each of the two microphone pairs not in line with each
other, Arrival time difference of sound pressure signal input to each microphone forming one of
the two microphone pairs and input to each microphone forming the other microphone pair And
estimating the arrival time difference of the pressure signal, the direction of the sound source
from the position coordinates and the speed of sound of the microphone. As described above,
even in the case of four microphones, if four microphones are arranged at each vertex of the
square, microphone distances L a and L b (L b = L a) different from the four microphones M1 to
M4 Since it is possible to configure two microphone pairs which are not in a straight line with
each other and have {square root over (2)}, the direction of the sound source can be accurately
estimated without lowering the upper limit of the measurable frequency.
03-05-2019
4
[0010]
The summary of the invention does not enumerate all necessary features of the present
invention, and a subcombination of these feature groups can also be an invention.
[0011]
FIG. 1 is a diagram showing a configuration of a sound source direction estimation apparatus
according to a first embodiment.
It is a figure which shows the structure of a sound extraction unit, and the example of
arrangement | positioning of a microphone. FIG. 7 is a diagram showing an arrangement example
of microphones according to a second embodiment. FIG. 18 is a diagram showing another
arrangement example of the microphones according to the third embodiment. It is a figure which
shows the example of arrangement | positioning of the conventional microphone.
[0012]
Hereinafter, embodiments of the present invention will be described based on the drawings.
Embodiment 1 FIG. 1 is a diagram showing the configuration of a sound source direction
estimation apparatus 1 according to the first embodiment, and the sound source direction
estimation apparatus 1 includes a sound collecting unit 10, a sound data input / output unit 21, a
storage unit 22, and a sound The pressure signal extraction means 23 and the sound source
direction estimation means 24 are provided. Each means from the storage means 22 to the sound
source direction estimation means 24 is constituted of, for example, software of a personal
computer and a memory. The sound collecting unit 10 includes a sound collecting unit 11 having
five microphones M1 to M5, a square plate-like flat plate 12 on which the microphones M1 to
M5 are mounted, and a support member 13 for supporting the flat plate 12. A base 14 on which
the support member 13 is installed, and a temperature sensor 15 installed on the base 14 are
provided. Here, the horizontal direction in the figure is taken as the horizontal direction, and the
vertical direction is taken as the vertical direction. In this example, as shown in FIG. 2A, the first
microphone M1 and the third microphone M3 are disposed at a position separated by a
predetermined distance L L in the longitudinal direction, and the second microphone M2 and the
third microphone M2 are The fourth microphone M4 is disposed in the lateral direction at a
position separated by the same predetermined interval L L as above, and the fifth microphone
M5 is a line connecting the first microphone M1 and the third microphone M3. , An intersection
point of a line segment connecting the second microphone M2 and the fourth microphone M4
03-05-2019
5
(here, an internal dividing point internally dividing 1: 3 between the first microphone M1 and the
third microphone M3) Placed in As shown in FIG. 2B, the microphones M1 to M5 are mounted on
the flat plate 12 so that the vibrating membrane surface is at substantially the same position as
one surface of the flat plate 12 (hereinafter referred to as the front surface 12a). Here, the
microphones M1 to M5 are flat when the direction from the rear surface 12b of the flat plate 12
toward the front surface 12a is a front direction, as indicated by the rear surface 12b and the
arrow in FIG. The sound pressure signal of the sound transmitted from the front of the face plate
12 is detected.
[0013]
Hereinafter, the position of the fifth microphone M5 is the origin (0, 0, 0), the horizontal
direction of the plane plate 12 is the x axis direction, the vertical direction is the y axis direction,
and the direction perpendicular to the plane plate 12 is the z axis direction. (See FIG. 2 (a)). The
horizontal angle θ is measured on the y-axis as θ = 0 and clockwise from the y-axis, and the
elevation angle φ is measured on the xy plane as φ = 0. In this example, the coordinates of the
position of the first microphone M1 are (0, L S, 0), the coordinates of the position of the
microphone M 2 (L M, 0, 0), and the coordinates of the position of the microphone M 3 (0, −
The coordinates of L M, 0) and the position of the microphone M 4 are (−L S, 0, 0). The
relationship between L S, L M and L L is L L = L S + L M, and L S <L M <L L. Here, assuming that
the microphone pair (M1, M3) is a first microphone pair and the microphone pair (M2, M4) is a
second microphone pair, the arrangement of the microphones M1 to M5 in this example is the
same as that of the microphones M1 to M4. The square to be formed is a square in which two
diagonal lines are orthogonal to each other, and the second microphone pair (M2, M4) is a first
microphone pair (M1, M3) centered on the fifth microphone. , Is located at a position rotated 90
degrees counterclockwise.
[0014]
By arranging the microphones M1 to M5 as shown in FIG. 2A in this manner, the microphone
spacings L S, L M, and L L different from the five microphones M1 to M5 (L S <L M < It is
possible to construct three microphone pairs that are not in line with one another, with L L). That
is, the first and second microphone pairs (M1, M3), (M2, M4) are the microphone pairs having
the longest microphone distance L L, and the microphone pairs (M1, M5), (M5, M4) are The
microphone pairs having the shortest microphone spacing L S (hereinafter referred to as the fifth
and sixth microphone pairs). Also, the microphone pairs (M5, M3), (M2, M5) are microphone
pairs (hereinafter referred to as third and fourth microphone pairs) having an intermediate
03-05-2019
6
microphone distance L M.
[0015]
The sound data input / output means 21 includes an amplifier 21a and an A / D converter 21b.
The amplifier 21a includes a low pass filter, removes high frequency noise components from the
sound pressure signal of the sound sampled by the microphones M1 to M5, amplifies each sound
pressure signal, and outputs the amplified signal to the A / D converter 21b. The A / D converter
21b A / D converts the sound pressure signal, and sends the A / D converted sound pressure
signal to the storage unit 22 as sound pressure waveform data. The storage unit 22 stores sound
pressure waveform data for each of the microphones M1 to M5. The sound pressure signal
extraction unit 23 includes a low-pass extraction unit 23 a, a mid-range extraction unit 23 b, and
a high-pass extraction unit 23 c, and the sound pressure signal of the microphones M1 to M5
stored in the storage unit 22 Extract the sound pressure signal of the microphone pair. That is,
since the horizontal angle θ and the elevation angle φ, which are sound source directions, are
calculated for each frequency, it is not necessary to obtain the arrival time difference of the first
to fifth microphone pairs for all the calculated frequencies. That is, the low-range extraction unit
23a extracts sound pressure waveform data of the sound pressure signals of the microphones
M1, M2, M3, and M4 constituting the microphone pair having the longest microphone distance L
L to obtain the sound source direction estimation unit 24. Send to The mid-range extraction unit
23 b extracts sound pressure waveform data of the sound pressure signals of the microphones M
2, M 3 and M 5 constituting the microphone pair having an intermediate microphone distance L
M and sends it to the sound source direction estimation unit 24. The high-frequency extracting
unit 23c extracts sound pressure waveform data of the sound pressure signals of the
microphones M1, M4, and M5 constituting the microphone pair having the shortest microphone
distance L S and sends the sound pressure signal data to the sound source direction estimating
unit 24.
[0016]
The sound source direction estimation means 24 performs frequency analysis of the sound
pressure waveform data extracted by the sound pressure signal extraction means 23 by FFT, and
the arrival time difference D corresponding to the phase difference between the two
microphones (Mi, Mj) as a pair The direction of the sound source is calculated for each frequency
from the obtained arrival time difference D ij and the speed of sound c calculated using the
temperature measured by the temperature sensor 15 for each frequency. Specifically, the
horizontal angle θ and the elevation angle φ of the sound in the low frequency region (for
03-05-2019
7
example, a region where the frequency is 1700 Hz or less when L L = 100 mm and c = 340 m / s)
are determined by the low frequency extraction unit 23a. Arrival time difference D 13 between
the microphones M 1 and M 3 constituting the extracted first microphone pair (M 1, M 3) and
arrival time difference D between the microphones M 2 and M 4 constituting the second
microphone pair (M 2, M 4) It is calculated by the following equation [Equation 1] using 24, the
sound speed c, and the microphone interval L L. Here, the arrival time difference D ij is obtained
by obtaining the cross spectrum P ij (f) of the signal input to the two microphones Mi and M j as
a pair, and further, the phase angle information Ψ (rad) of the target frequency f Is calculated by
the following equation [Equation 2].
[0017]
Further, the horizontal angle θ and the elevation angle φ of the sound in the intermediate
frequency region (for example, 1700 Hz to 2266 Hz when L M = 75 mm) are the third
microphone pair (M 5, M 3) extracted by the mid-range extraction unit 23 b. Difference in arrival
time D 53 between the microphones M5 and M3 constituting the), arrival time difference D25
between the microphones M2 and M5 constituting the fourth microphone pair (M2, M5), the
speed of sound c, and the microphone spacing L M And are calculated by the following equation
[Equation 3]. Similarly, the horizontal angle θ and the elevation angle φ of the sound in the high
frequency range (for example, 2266 Hz to 6800 Hz when L S = 25 mm) is the fifth microphone
pair (M1, M5) extracted by the high-frequency extracting unit 23c. Difference in arrival time D
15 between the microphones M 1 and M 5 constituting the), arrival time difference D 54
between the microphones M 5 and M 4 constituting the sixth microphone pair (M 5, M 4), sound
speed c, and microphone spacing L S And are calculated by the following equation [Equation 4].
[0018]
As described above, in the first embodiment, the first to fourth microphones M1 to M4, which are
not on a straight line with each other, are arranged on the flat plate 12, and the positions of the
first to fourth microphones M1 to M4 are set. The fifth microphones M5 are arranged at the
intersections of diagonals of the quadrangle as apexes, and different microphone intervals L S, L
M, and L L selected from these five microphones M1 to M5 (L S <L M <L L ) For each of the two
microphone pairs which are not in line with each other, the arrival time difference of the sound
pressure signal input to each of the microphones constituting one of the two microphone pairs,
and the other microphone pair From the arrival time difference of the sound pressure signal
input to each of the constituent microphones, the position coordinates of the microphones M1 to
M5, and the sound speed c Since the direction of the source was estimated, for the direction of
the sound source generating low frequency sound, the data of the microphone pair (L = L L)
microphone pair (M1, M3), (M2, M4) are used The sound source direction that generates high-
03-05-2019
8
frequency sound is estimated using data of microphone pairs (L = L S) microphone pair (M1, M5),
(M5, M4), and intermediate frequency The upper limit of the frequency that can be measured is
estimated using the data of the microphone pairs (M5, M3) and (M2, M5) with an intermediate
microphone spacing (L = L M) for the direction of the sound source that generates The direction
of the sound source can be accurately estimated without lowering it.
[0019]
Second Embodiment
In the first embodiment, the distance between the first microphone M1 and the third microphone
M3 is L L = 100 mm, and the fifth microphone M5 is located between the first microphone M1
and the third microphone M3. The internal division point internally dividing into 1: 3 is disposed,
but the microphone interval L L and the position of the fifth microphone M5 may be
appropriately determined according to the frequency range to be measured. FIG. 3 shows an
example in which the fifth microphone M5 is disposed at the middle point between the first
microphone M1 and the third microphone M3. In this case, the quadrangle formed by the
microphones M1 to M4 is a square, and The fifth microphone M5 is located at the center of the
square. Also in this case, the microphone pair having the longest microphone distance L L is the
first and second microphone pairs (M1, M3) and (M2, M4) as in the first embodiment. On the
other hand, the microphone pair having the shortest microphone distance (L = L L / 2) is the
microphone pair (M1, M5), (M2, M5) in addition to the microphone pair (M1, M5), (M5, M4). ,
Microphone pairs (M5, M3), (M5, M4), and microphone pairs (M5, M3), (M2, M5). Also,
microphone pairs having an intermediate microphone spacing (L = L L / √2) include microphone
pairs (M1, M4), (M2, M1) and microphone pairs (M1, M4), (M3, M4) , Microphone pairs (M2,
M3), (M2, M1), and microphone pairs (M2, M3), (M3, M4).
[0020]
Also in this case, the sound source direction estimation means 24 analyzes the frequency of the
sound pressure waveform data extracted by the sound pressure signal extraction means 23 by
FFT, and corresponds to the phase difference between the two microphones (Mi, Mj) as a pair.
The arrival time difference D ij is calculated for each frequency, and the direction of the sound
source is calculated for each frequency from the calculated arrival time difference D ij and the
speed of sound c calculated using the temperature measured by the temperature sensor 15. The
horizontal angle θ and the elevation angle φ of the sound in the low frequency region (for
03-05-2019
9
example, the region where the frequency is 3400 Hz or less when L L = 50 mm and c = 340 m /
s) are the same as in the first embodiment. Calculated by 5]. The horizontal angle θ and
elevation angle φ of the sound in the intermediate frequency range (3400 Hz to 4808 Hz when
L L = 50 mm) are the microphone pairs (M1, M4), (M2, M1) and the microphone pairs (M1, M4) ,
(M3, M4), microphone pair (M2, M3), (M2, M1), microphone pair (M2, M3), (M3, M4),
microphone pair (Ma, Mb), (Mc, Md) And the arrival time difference D ab between the
microphones Ma and Mb constituting the microphone pair (Ma, Mb), the arrival time difference D
cd between the microphones Mc and Md constituting the microphone pair (Mc, Md), and the
sound speed c. Using the microphone distance L L, the following equation [6] is calculated.
Similarly, the horizontal angle θ and the elevation angle φ of the sound in the high frequency
range (for example, 4808 Hz to 6800 Hz when L L = 50 mm) are set to the microphone pair in
addition to the microphone pair (M1, M5), (M5, M4) (M1, M5), (M2, M5), microphone pair (M5,
M3), (M5, M4), and microphone pair (M5, M3), (M2, M5), microphone pair (Mp, Mq) And (Mr,
Ms), the arrival time difference D pq between the microphones Mp, Mq constituting the
microphone pair (Mp, Mq) and the arrival between the microphones Mr, Ms constituting the
microphone pair (Mr, Ms) Using the time difference D rs, the sound velocity c, and the
microphone interval L L, the following equation [7] is calculated.
[0021]
As described above, if the square formed by the microphones M1 to M4 is a square and the fifth
microphone M5 is positioned at the center of the square, the horizontal angle θ and elevation
angle φ of the sound in the intermediate frequency range and the high frequency range The
horizontal angle θ and the elevation angle φ of the sound of can be determined four each.
Therefore, by averaging the horizontal angle θ and the elevation angle φ, the estimation
accuracy of the horizontal angle θ and the elevation angle φ of the sound in the intermediate
frequency region and the high frequency region can be greatly improved. As described above, by
disposing the microphones M1 to M5 at the apex and the center of the square, it is possible to
estimate the direction of the sound source more accurately without lowering the upper limit of
the measurable frequency.
[0022]
In the first and second embodiments, the microphones M1 to M4 are disposed at the vertexes of
a square whose diagonals are orthogonal to each other. However, the microphones M1 to M4
may be quadrilaterals whose diagonals are not orthogonal to each other. However, in this case,
since the calculation of the horizontal angle θ and the elevation angle φ becomes complicated,
03-05-2019
10
it is preferable to place the microphones M1 to M4 at the vertexes of a square whose diagonals
are orthogonal.
[0023]
Third Embodiment In the first and second embodiments, five microphones M1 to M5 are used to
form three microphone pairs which are different from each other and have different microphone
intervals L S, L M and L L. As shown, even with four microphones M1 to M4 arranged at the top
of a square with L m on one side, they have different microphone spacings (L 1 = √2 L m, L 2 =
L m), and are in line with each other Not two microphone pairs can be configured. The
microphone pairs (M1, M3), (M2, M4) are the microphone pairs having the longer microphone
distance L 1, the microphone pairs (M1, M4), (M2, M1) and the microphone pairs (M1, M4), (M3,
M4), the microphone pairs (M2, M3), (M2, M1), and the microphone pairs (M2, M3), (M3, M4)
are microphone pairs having a shorter microphone distance L 2 . The method of calculating the
horizontal angle θ and the elevation angle φ from the sound pressure signal of the sound input
to the microphones M1 to M4 is the same as in the case of the configuration in which the
microphones M1 to M5 are arranged at the center and the square described above. So I will omit
the explanation. Similarly, if n (n is an even number of 4 or more) microphones M1 to Mn are
arranged at each vertex of the n polygon, three or more microphones having different
microphone intervals but not on a straight line with each other Pairs can be configured.
[0024]
As mentioned above, although this invention was demonstrated using embodiment, the technical
scope of this invention is not limited to the range as described in the said embodiment. It is
obvious to those skilled in the art that various changes or modifications can be added to the
above embodiment. It is also apparent from the scope of the claims that the embodiments added
with such alterations or improvements can be included in the technical scope of the present
invention.
[0025]
DESCRIPTION OF SYMBOLS 1 sound source direction estimation device, 10 sound collecting unit,
11 sound collecting means, 12 flat plate, 12 a flat plate front surface, 12 b flat plate rear surface,
13 support member, 14 base, 15 temperature sensor, 21 sound data input / output means , 21a
amplifier, 21b A / D converter, 22 storage means, 23 sound pressure signal extraction means,
23a low band extraction part, 23b mid band extraction part, 23c high band extraction part, 24
03-05-2019
11
sound source direction estimation means, M1 ~ M5 Microphone.
03-05-2019
12
Документ
Категория
Без категории
Просмотров
0
Размер файла
24 Кб
Теги
jp2016127430
1/--страниц
Пожаловаться на содержимое документа