close

Вход

Забыли?

вход по аккаунту

JPH06292293

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 JPH06292293
[0001]
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a
so-called superdirective microphone device using an adaptive noise canceller.
[0002]
2. Description of the Related Art For example, in a camera integrated VTR, while photographing
an object, sounds around the object are simultaneously recorded. In this case, it is generally
considered to pick up only the sound from the direction of the subject. That is, for example, a
microphone device having a directional characteristic such that only sound from the front of the
camera is picked up is used.
[0003]
As an example of this type of microphone device, for example, a so-called gun microphone is
known. It comprises a pipe section which extends to the front of the diaphragm. Further, a large
number of through holes are provided on the side wall of this pipe portion, and it has directivity
having high sensitivity to sound from the front (in the direction opposite to the diaphragm) of the
pipe portion in the centerline direction. Is configured as.
03-05-2019
1
[0004]
However, this microphone requires a long pipe portion and has the drawback of becoming large.
Moreover, it is uni-directional with high sensitivity only in front of the microphone, and only
fixed directivity can be obtained. Therefore, it is difficult to cope with the case where it is desired
to pick up not only the sound from the desired sound arrival direction but also the sound from
the side around the camera, for example, and there is no freedom in the direction of directivity.
[0005]
Therefore, the applicant has proposed a microphone device which can be made compact and
have superdirective characteristics by applying an adaptive noise canceller as Japanese Patent
Application No. 4-143209.
[0006]
FIG. 9 shows the basic configuration. First, an adaptive noise canceller will be described.
[0007]
In FIG. 9, 10 is an adaptive noise canceller, 1 is its main input terminal, 2 is its reference input
terminal, and the main input signal inputted through the main input terminal 1 is sent to the
synthesis circuit 4 through the delay circuit 3. Supplied.
Further, the reference input signal input through the reference input terminal 2 is supplied to the
synthesis circuit 4 through the adaptive filter circuit 5 and is subtracted from the signal from the
delay circuit 3.
The output of the synthesis circuit 4 is fed back to the adaptive filter circuit 5 and is led to the
output terminal 6.
[0008]
In this adaptive noise canceller, the main input signal is the sum of the desired signal s and the
03-05-2019
2
noise signal n0 uncorrelated with it. On the other hand, the noise signal n1 is input as the
reference input signal. The noise signal n1 of the reference input is uncorrelated with the desired
signal s, but is correlated with the noise signal n0.
[0009]
The adaptive filter circuit 5 filters the reference input noise signal n1 and outputs a signal y
which approximates the noise signal n0. In this case, the adaptive filter circuit 5 updates the
filtering coefficient of the reference input noise signal n1 by the predetermined adaptive
algorithm so that the subtraction output (residual output) e of the combining circuit 4 is
minimized. Go on.
[0010]
As the output signal y of the adaptive filter circuit 5, it is also possible to obtain a signal having
the same amplitude as that of the noise signal n0. The delay circuit 3 compensates for the time
delay required for the arithmetic processing in the adaptive filter circuit 5, the propagation time
in the adaptive filter, and the like, and is used to time align with the signal to be subtracted.
[0011]
The principle of the adaptive noise canceller will be described below.
[0012]
Now, assuming that the desired signal s, the noise n0, the noise n1, and the output signal y are
statistically stationary and the average value is zero, the residual output e becomes e = s + n0−y.
The expected value of this squared is E [e2] = E [s2] + E [(n0-y) 2] + 2E [s (n0-y) because the
desired signal s has no correlation with the noise n0 and the output y. ]] = E [s2] + E [(n0-y) 2].
Assuming that the adaptive filter circuit 5 converges, the adaptive filter circuit 5 updates the
adaptive filter coefficients such that E [e2] is minimized. At this time, E [s2] is not affected, so
Emin [e2] = E [s2] + Emin [(n0-y) 2].
03-05-2019
3
[0013]
That is, E [(n0-y) 2] is minimized by minimizing E [e2], and the output y of the adaptive filter
circuit 5 becomes an estimate of the noise signal n0. The expected value of the output from the
combining circuit 4 is only the desired signal s. That is, adjusting the adaptive filter circuit 5 to
minimize the total output power is equal to the subtraction output e being the least squares
estimated value of the desired speech signal s.
[0014]
Although the output e generally has some noise remaining in the signal s, since the output noise
is given by (n 0 −y), minimizing E [(n 0 −y) 2] is an output It is equivalent to maximizing the
signal to noise ratio.
[0015]
In some cases, the synthesis circuit 4 may be acoustic synthesis means.
That is, the adaptive filter circuit 5 forms a noise cancellation signal -y having the same
amplitude as that of the noise and the opposite phase, and supplies this to a speaker or the like
to acoustically add to the main signal to reduce the noise. . The residual e in this case is picked
up by the residual detection microphone.
[0016]
The adaptive filter circuit 5 can be realized either by an analog signal processing circuit or a
digital signal processing circuit, but in general, a digital processing circuit using a DSP (digital
signal processor) It is made up of
[0017]
An example of the configuration of the adaptive filter circuit 5 in the case of the configuration of
the digital processing circuit is shown in FIG.
03-05-2019
4
[0018]
In this example, the adaptive filter circuit 5 comprises an FIR filter type adaptive linear combiner
100 and a filter coefficient update computing means 110.
The adaptive filter circuit 5 can be configured as software by a DSP on which a microcomputer is
mounted.
In this example, the algorithm for updating the filter coefficient will be described as using LMS
(Least Mean Squares), which is frequently used because the amount of calculation is small and
practical.
[0019]
The LMS method will be described with reference to FIG. As shown in FIG. 11, in this case, an FIR
filter type adaptive linear combiner 100 is used. The adaptive linear coupler 100 includes a
plurality of delay circuits DL1, DL2,... DLm (m is a positive integer) each having a delay time Z-1
of unit sampling time, an input noise n1, and each delay circuit DL1. , DL2,..., A weighting circuit
MX0, MX1, MX2,... MXm for multiplying the output signal of the DLm and the weighting
coefficient (filter coefficient), and an adding circuit 101 for adding the outputs of the weighting
circuits MX0 to MXm. The output of the adder circuit 101 is the signal y described in FIG.
[0020]
The weighting coefficients to be supplied to the weighting circuits MX0 to MXm are formed by
the filter coefficient arithmetic circuit 110 based on the residual signal e from the synthesizing
circuit 4 and the reference input n1 by the LMS algorithm. The algorithm executed by the filter
coefficient calculation circuit 110 is as follows.
[0021]
Now, as also shown in FIG. 11, the input vector Xk at time k is Xk = [x0k x1k x2k ... xmk] T, the
output is yk, and the weighting coefficient is wjk (j = 0, 1, 2, ... Assuming that m), the input /
output relationship is as shown in the following equation 1:
03-05-2019
5
[0023]
Then, if the weight vector Wk at time k is defined as Wk = [w0k w1k w2k... Wmk] T, the input /
output relation is given by yk = Xk T · Wk (1).
Here, assuming that the desired response is dk, the residual ek is ek = dk −yk = dk −Xk T · Wk
(2)
[0024]
In the LMS method, updating of the weight vector is sequentially performed according to the
equation (3) as Wk + 1 = Wk + 2 μ · ek · Xk (3). The initial value of the weighting factor is set to
a constant value or a random value. Here, μ is a gain factor (step gain) that determines the speed
and stability of adaptation.
[0025]
In the above equation (3), the vector for correcting the coefficient vector Wk at a certain time k is
the second term of the right side of equation (3), but the gain factor μ and the instantaneous
error ek are both scalar values, Directly influence the value. Since the reference input vector Xk
also works in the form of a product, this also influences the correction value. The average
convergence time constant τa is represented by τa = (n + 1) / 4μ · trE [Xi Xj T]. Here, n is the
order of the reference input vector (corresponding to the number of taps of the FIR filter), and
trE [Xi Xj T] is the average power of the reference input. That is, the larger the number of taps of
the FIR filter, the slower the convergence speed, and the larger the gain factor μ, the faster the
convergence speed.
[0026]
In the case of a stationary signal, if the convergence speed is fast, the final residual noise level is
large, and if the convergence is slow, the final noise level is small. However, when the target
03-05-2019
6
signal fluctuates like speech, the nature of the signal changes before convergence is complete, so
the amount of cancellation becomes larger if the convergence speed is faster to some extent.
[0027]
The value of the gain factor μ is such that the following condition is satisfied so that the output
y of the adaptive filter circuit 5 approaches that canceling noise n 0 and the output of the device
converges to be equal to the desired signal s You need to be satisfied. If 0 <μ <(power of signal)
/ (number of taps of FIR filter + 1) (4) becomes larger than the range of equation (4), the output y
of the adaptive filter circuit 5 diverges, and It produces a loud noise as an output.
[0028]
Conventionally, the value of the gain factor μ is determined by the above equation (1) taking
into account the number of taps of the adaptive filter (filter order) directly affecting the quality of
the processed signal (filter order) and the magnitude (power) of the reference input signal. In
order to satisfy 4) and allow the adaptive filter circuit 5 to operate normally (converge), it is set
to a constant value. Generally, it is calculated from the maximum power of the reference input to
compensate for the stability of the system.
[0029]
The microphone device proposed above is configured using the adaptive noise canceller as
described above. That is, as shown in FIG. 9, the first microphone 7 for main input sound
collection and the second microphone 8 for reference input sound collection are provided, and
the output signal of the first microphone 7 is input to the main input terminal 1. And the output
signal of the second microphone 8 is input to the reference input terminal 2.
[0030]
In this case, as shown in FIG. 10, when the desired sound arrival direction is the direction of the
arrow AR, the main input microphone 7 uses a nondirectional microphone. Alternatively, a
unidirectional microphone is placed with its main pointing axis in the desired sound arrival
03-05-2019
7
direction.
[0031]
In addition, as shown in FIG. 10, the microphone for reference input 8 has sensitivity in the
incident direction of the unnecessary sound as noise to be eliminated, and the desired sound
arrival direction AR is the direction with the minimum sensitivity, as shown in FIG. Arrange and
use it as If the noise arrival direction to be excluded is 90 degrees to the desired sound arrival
direction, use a bi-directional microphone so that the desired sound arrival direction AR is the
directivity null direction You may
[0032]
With such a configuration, the reference input signal contains almost no component of the
desired sound, so the reference input signal becomes a signal uncorrelated with the desired
sound, and an unnecessary signal (noise component) in the main input. And a signal having a
correlation with Therefore, if the second microphone 8 for collecting the reference input is
configured to have a predetermined sensitivity in the arrival direction of the unnecessary signal
to be reduced, the unnecessary signal component included in the main input is adaptively
canceled. , And only the desired sound signal can be obtained at the output terminal.
[0033]
As described above, by using the adaptive noise canceller, it is possible to realize a superdirective
microphone device having a sharp directivity in the direction of the sensitivity minimum of the
unidirectional microphone 8 for reference input.
[0034]
However, in the above-described microphone device proposed earlier, it has been found that the
amount of noise reduction differs depending on the direction of arrival of sound.
In particular, in the case of non-stationary signals such as human voice signals, there is a
problem that a large reduction amount can not be obtained.
03-05-2019
8
[0035]
In view of the above, it is an object of the present invention to provide a microphone device
capable of obtaining an appropriate amount of noise reduction regardless of the direction of
arrival of sound.
[0036]
SUMMARY OF THE INVENTION The inventors of the present invention have made it clear that
the problem described above is that the step gain .mu. For updating the coefficients of the
adaptive filter circuit of the adaptive noise canceller is constant regardless of the direction of
arrival of the sound. I found it to be.
[0037]
FIG. 5 shows the sound (for example, human voice) incident from another direction when the
direction (the direction of the highest sensitivity) of the directional axis of the previously
proposed microphone device is directed to the front direction (the 0 degree direction). It is
experimental data of the relationship between the amount of reduction and the step gain
parameter.
This is an experimental result in the case where the incident direction of the sound is 90 degrees
with respect to the front direction and the case where it is 180 degrees (rear direction).
[0038]
As is clear from FIG. 5, it was found that the value of the step gain μ at which the maximum
reduction amount can be obtained is different depending on the sound incident direction.
Therefore, when the adaptive processing is performed with a constant step gain parameter
regardless of the direction of arrival of sound, the adaptation speed varies depending on the
direction of arrival of sound, and the amount of noise reduction varies depending on the
direction of arrival of sound.
03-05-2019
9
[0039]
The present invention is characterized in that an optimum step gain parameter is set according to
the arrival direction (incident direction) of the sound in consideration of the above points.
[0040]
That is, in the microphone device according to the present invention, the first microphones M1
and M3 for picking up the desired sound, and the directivity having low sensitivity in the
direction of arrival of the desired sound, correspond to the reference numerals of the
embodiments described later. A second microphone M3, adaptive filter means 15 to which the
output signal of the second microphone is supplied, and combining means 14 for subtracting the
output signal of the adaptive filter means from the output signal of the first microphone ,
Detecting means 21 for detecting the incident direction of sound, and the adaptive filter means
15 adjusts its output signal so that the output power of the synthesizing means 14 is minimized.
The gain factor parameter μ of the unit 110 is characterized in that it is controlled according to
the incident direction of the sound detected by the detection means.
[0041]
In the present invention having the above-described configuration, the gain factor (step gain)
parameter of the adaptive operation unit of the adaptive filter means is the most in that direction
according to the direction detected by the detection means 21 of the incident direction of sound.
It is selected to a suitable value.
[0042]
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT An embodiment of the microphone
device according to the present invention will be described below with reference to FIG.
This example is an example where the adaptive filter circuit of the adaptive noise canceller has a
digital configuration.
[0043]
In FIG. 1, M1, M2 and M3 are first, second and third microphones, and in this example, they are
03-05-2019
10
both made up of unidirectional microphones.
The arrangement of the microphones M1 to M3 is shown in FIG. 2, and the directivity pattern
(polar pattern) of the unidirectional microphones used as the microphones M1 to M3 is shown in
FIG.
[0044]
As shown in FIG. 3, when the maximum sensitivity is "1", this unidirectional microphone has a
sensitivity of "0.5" in the direction of ± 90 degrees with respect to the direction of the maximum
sensitivity, The sensitivity in the direction of 180 degrees with respect to the direction is 0 .
[0045]
Then, assuming that the desired sound arrival direction is the 0 degree direction in FIG. 2, the
first microphone M1 has its main directional axis direction (the maximum sensitivity direction
indicated by the arrow superimposed on the directional pattern in FIG. 2; the same applies
hereinafter) Of the second microphone M2 is oriented 90 degrees in the direction of its main
directional axis, and the third microphone M3 is oriented 180 degrees in the direction of its main
directional axis. Are arranged.
That is, the first to third microphones M1 to M3 are disposed in directions in which the main
directional axis directions are orthogonal to each other.
[0046]
The output signal of the first microphone M1 and the output signal of the third microphone M3
are supplied to the adder circuit 11 and added.
The output signal of the addition circuit 11 is non-directional because it is an addition signal of a
microphone whose direction of the main directional axis is 180 degrees different from each
other. This is equal to the output characteristic of the main input microphone 7 of FIG. 9
described above. An output signal of the addition circuit 11 is converted into a digital signal by
an A / D converter 12 and supplied to a subtraction circuit 14.
03-05-2019
11
[0047]
The third microphone M3 corresponds to the reference input microphone 8 described above. The
output of the third microphone M 3 is converted into a digital signal by the A / D converter 13
and supplied to the adaptive filter circuit 15. The adaptive filter circuit 15 includes an adaptive
linear combiner 100 and a filter coefficient calculation circuit 110 using, for example, the LMS
method as an algorithm for updating filter coefficients, as shown in FIG.
[0048]
Then, the output signal of the adaptive linear coupler 100 is supplied to the subtraction circuit
14 and is subtracted from the signal from the A / D converter 12. The output signal of the
subtraction circuit 14 is field-backed to the filter coefficient calculation circuit 110 of the
adaptive filter circuit 15, and is converted back to an analog signal by the D / A converter 16 and
is derived to the output terminal 17.
[0049]
The subtraction circuit 14 and the adaptive filter circuit 15 constitute an adaptive noise
canceller, and the filter coefficient calculation circuit 110 updates the filter coefficient according
to the equation (3). However, in this case, as described below, the step gain μ of the update
equation of the equation (3) is controlled to be set to an optimal value according to the signal
arrival direction.
[0050]
That is, the output signals of the first to third microphones M1 to M3 are supplied to the
direction detection and parameter control circuit 21.
[0051]
When three unidirectional microphones M1 to M3 are arranged as shown in FIG. 2, for these
03-05-2019
12
three microphones M1 to M3, sensitivity matrices as shown in FIG. be able to.
Generally, the sensitivity of the microphone is proportional to the output signal level, so referring
to FIG. 2, the incident direction of the sound can be determined from the magnitude relationship
between the output signal levels of these three microphones M1 to M3. Therefore, the control
circuit 21 detects the incident direction of the sound by comparing the levels of the output
signals of the three microphones M1 to M3.
[0052]
Further, in the control circuit 21, a correspondence table of the sound incident direction and the
value of the optimum step gain μ is created in advance by experiment or the like, and is stored
in the memory. The control circuit 21 refers to the correspondence table to determine the value
of the optimum step gain for the incident direction of the detected sound. Then, the value of the
optimum step gain is supplied to the filter coefficient calculation circuit 110 as a parameter.
[0053]
Therefore, the adaptive filter circuit 15 always performs the adaptive process of updating the
filter coefficient with the value of the step gain optimum for the incident direction of the sound at
that time, and as a result, both the adaptation speed and the noise reduction amount are better. A
superdirective microphone device can be realized.
[0054]
In the example of FIG. 1, all the outputs of the three microphones M1 to M3 are used to detect
the incident direction of sound, but as is apparent from FIG. 4, the microphone M1 and the
microphone M2 or the microphone M2 and The detection of the incident direction of sound can
be performed from the outputs of the two microphones in combination with the microphone M3.
Further, in this case, the two microphones do not have to point the direction of the main
directional axis in the direction orthogonal to each other, and if the directions of the main
directional axes cross each other, the sound incident direction Can be detected.
03-05-2019
13
[0055]
Next, FIG. 6 is a block diagram of another embodiment of the microphone device according to the
present invention. In this example, using three nondirectional microphones M11, M12 and M13,
so-called two-dimensional acoustic intensity is calculated based on the outputs of these three
microphones M11, M12 and M13, and the incident direction of sound is obtained. Is an example
in the case of detecting.
[0056]
The sound intensity is defined by the amount of energy passing a unit area perpendicular to the
traveling direction of sound per unit time, and is represented by (sound pressure) × (particle
velocity) [W / m 2]. The measurement of the sound intensity can be performed using two
microphones arranged in parallel with the traveling direction of the sound, and the particle
velocity of the sound is approximated from the difference between the sound pressure values,
The sound pressure at the midpoint between two microphones can be estimated.
[0057]
Therefore, for example, if two microphones are arranged in a direction orthogonal to each other
and two-dimensional sound intensity is measured, detecting the arrival direction of sound by
synthesizing the sound intensity vectors in the respective two microphone array directions. Can.
[0058]
In this example, based on the above points, as shown in FIG. 7, the microphone M11 and the
microphone M12 are disposed apart by a predetermined distance d in the 90 degree direction (x
direction) and the 0 degree direction ( The microphones M11 and M13 are disposed apart by the
same distance d in the y direction).
Then, output signals of these three microphones M11, M12, and M13 are supplied to the
direction detection and parameter control circuit 22.
[0059]
03-05-2019
14
Then, in the control circuit 22, an acoustic intensity vector in the x direction is determined from
the outputs of the microphones M11 and M12, and an acoustic intensity vector in the y direction
is determined from the outputs of the microphones M11 and M13. Then, from these twodimensional acoustic intensity vectors, the flow of the energy of the sound at that moment, that
is, the arrival direction of the sound is detected.
[0060]
Also in this example, the control circuit 22 has a memory in which a correspondence table of the
incident direction of sound and the value of the optimum step gain μ is stored, and referring to
the correspondence table, the control circuit 22 Determine the value of the optimum step gain
for the incident direction. Then, the value of the determined optimum step gain is supplied to the
filter coefficient calculation circuit 110 as a parameter.
[0061]
Further, in the case of this example, the output signal of the microphone M11 is supplied to the A
/ D converter 11, and the digital output signal thereof is supplied to the subtraction circuit 14 as
the main input. Further, the output signal of the microphone M13 is subtracted from the output
signal of the microphone M11 by the subtraction circuit 18, and the direction of the highest
sensitivity from the subtraction circuit 18 is 180 degrees directed opposite to the desired sound
direction. A signal is obtained. The output signal of the subtraction circuit 18 is supplied to an A
/ D converter 13, and the output digital signal is supplied to the adaptive filter circuit 15 as a
reference input. Then, the adaptive processing is performed in the state of the optimal step gain
μ according to the incident direction of the sound described above, and the noise reduction
processing is performed.
[0062]
Incidentally, as shown in FIG. 8, by arranging the bidirectional microphone M21 and the
microphone M22 so that the directions of the main directivity axes thereof are different from
each other by 90 degrees, in the same manner as described above, two-dimensional Since the
sound intensity vector can be obtained, the output direction of the sound may be detected using
the outputs of these two microphones M21 and M22.
03-05-2019
15
[0063]
The directions of the two-dimensional sound intensity vectors do not have to be orthogonal to
each other, and may be directions intersecting each other.
[0064]
It is needless to say that the algorithm for updating the filter coefficients is not limited to the
LMS method described above, and for example, a learning algorithm and other algorithms can be
used.
[0065]
As described above, according to the present invention, in the superdirective microphone device
using the adaptive noise canceller, in consideration of the fact that the optimal step gain differs
depending on the direction of arrival of the sound, the direction of arrival of the sound is Since
the optimum step gain is set in accordance with the direction of arrival of the detected sound, it
is possible to realize a superdirective microphone excellent in both the adaptation speed and the
noise reduction amount.
03-05-2019
16
1/--страниц
Пожаловаться на содержимое документа