close

Вход

Забыли?

вход по аккаунту

?

DESCRIPTION JPH0522788

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 JPH0522788
[0001]
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a
noise reduction apparatus for reducing and extracting noise in a main input such as a voice
signal mixed with noise.
[0002]
2. Description of the Related Art In a communication apparatus or a recording apparatus having
a microphone for picking up an audio signal such as voice or music in general, noise mixed in the
microphone is suppressed to make it easy to hear received voice or reproduced voice. In order to
do this, a noise reduction device is used. The noise reduction device may be applied to a voice
recognition device and used for the purpose of reducing false recognition due to noise mixing.
[0003]
As this noise reduction device (noise canceller), an adaptive noise reduction device using an
adaptive filter is known in which filter characteristics are adaptively controlled according to an
input signal. In this adaptive noise reduction device, the main input such as a voice signal
containing noise obtained by receiving (picking up) voice etc. with a microphone etc. is sent to
the subtractor, and only noise is received with another microphone etc ( The reference input
obtained as a result of sound collection is filtered by an adaptive filter, and the filter output is
10-04-2019
1
sent to the subtractor to be subtracted from the main input. This adaptive filter is for modifying
and controlling internal filter coefficients and the like so as to minimize the power of the output
(so-called residual) from the subtractor.
[0004]
By the way, an FIR (finite impulse response) filter, for example, is used as the above-mentioned
adaptive filter, and this FIR filter is a path from a noise pickup microphone to a sound pickup
microphone such as voice. The transfer characteristic of is approximated as a linear
characteristic. Therefore, although the approximation of the transfer characteristic is correctly
performed, the noise will be completely eliminated, but in practice, the approximation can often
only be incompletely performed. As described above, if the approximation of the transfer
characteristic by the adaptive filter is not perfect in the adaptive noise reduction device, a
residual noise component correlated with the noise component will remain in the residual.
[0005]
One of the reasons why the above approximation can not be accurately performed is that the
circuit scale of the FIR filter, in particular, the number of taps can not be made sufficiently large.
Simply increasing the number of taps increases the accuracy of the approximation, but it also
takes a long time to converge the learning when modifying the filter coefficients. Here, the larger
the reference input, the higher the convergence speed, but the larger the error of the above
approximation, and the smaller the reference input, the smaller the error but the convergence
speed.
[0006]
The present invention has been made in view of such circumstances, and can reduce the adverse
effect due to residual noise included in the noise reduction output, and can perform noise
reduction operation optimum for the input signal according to the state. The purpose is to
provide a reduction device.
[0007]
According to the present invention, there is provided a noise reduction apparatus comprising:
signal level detection means for detecting the signal level of a main input in which noise is mixed
in a signal component; subtraction means to which the main input is supplied; The filter
10-04-2019
2
characteristic of the filtering process is controlled to send the filtered output to the subtracting
means based on the reference input of the noise component and to subtract it from the main
input to minimize the power of the output from the subtracting means. The above problem is
solved by including an adaptive filter and a control means for controlling the processing speed of
the adaptive filter in accordance with the detection output from the signal level detection means.
[0008]
Here, as the control means, for example, the level of the reference input is controlled according
to the detection output from the signal level detection means and supplied to the adaptive filter,
or the output from the subtraction means is the signal level The configuration may be such that
the level control is performed according to the detection output from the detection means to
return to the above-mentioned adaptive filter.
[0009]
The processing speed of the adaptive filter is controlled in accordance with the signal level of the
main input.
This makes it possible to increase or decrease the degree of adaptive noise reduction processing
according to how the noise is heard in the auditory sense, increase the noise reduction when the
noise is heard on the ear, and reduce the noise reduction when the signal level is large and the
noise can not be heard. Generation of distortion due to processing.
[0010]
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First, FIG. 1 is a block circuit
diagram showing a schematic configuration of a noise reduction apparatus according to an
embodiment of the present invention.
In FIG. 1, a main input (s + na) in which a noise component na is mixed with a signal component s
such as voice is supplied to the input terminal 11, and the main input serves as a subtracting
means via the terminal 12 Sent to the container 15.
10-04-2019
3
A noise component nb obtained by, for example, collecting noise sound from a noise source with
a microphone is supplied to the input terminal 13, and in this embodiment, it is used as a
reference input x via the gain control circuit 19. It is sent to the reference input terminal 16 a of
the adaptive filter 16. The output y from the adaptive filter 16 is sent as a subtraction signal to
the adder 15 serving as the subtraction means, and is subtracted from the main input. The output
from the adder 15, the so-called residual e, is taken out at the output terminal 17 and is also
returned via the terminal 16c of the adaptive filter 16. The adaptive filter 16 adaptively controls
the filter coefficients and the like so that the power of the residual e is minimized. Here, although
the signal component s and the noise components na and nb are uncorrelated, the noise
components na and nb are correlated.
[0011]
Further, the signal level of the main input supplied to the input terminal 11 is detected by the
signal level detection circuit 18, and the detected output is sent to the gain control circuit 19 as a
gain control signal. Here, for example, a root mean square (RMS) / dB (decibel) converter or the
like is used as the signal level detection circuit 18, and the gain (gain) of the gain control circuit
(gain controller) 19 is controlled by its output voltage. Just do it. The gain control circuit 19
controls the gain g in accordance with the signal level of the main input, and the noise
component nb supplied to the input terminal 13 is amplified by the controlled gain g of the gain
control circuit 19 and the above reference It becomes an input x (x = gnb) and is sent to the
terminal 16a of the adaptive filter 16.
[0012]
The input terminals 12 and 16a to the output terminal 17 in FIG. 1 correspond to the
configuration of a normal adaptive noise reduction circuit (adaptive noise canceller), and digital
processing is performed. If the signals supplied to the input terminals 11 and 13 in FIG. 1 are
analog signals, A / D conversion (analog / digital conversion) may be performed at the positions
of the input terminals 12 and 16a. In addition, A / D conversion may be performed at positions
11 and 13, respectively. In this case, it is necessary to use circuits which perform level detection
and gain control digitally as the signal level detection circuit 18 and the gain control circuit 19,
respectively. Furthermore, when it is desired to obtain the noise reduction output as an analog
signal, D / A conversion (digital / analog conversion) may be performed at the position of the
output terminal 17.
10-04-2019
4
[0013]
By the way, the adaptive filter 16 generates a pseudo output (pseudo noise) y of the noise
component na by learning based on the reference input x, and as shown in FIG. And 22. The
reference input x supplied via the reference input terminal 16 a is supplied to the filter unit 21
and the adaptive algorithm unit 22. The output y from the filter unit 21 is taken out as the
output of the adaptive filter 16 through the terminal 16b, and the residual e supplied through the
terminal 16c is supplied to the adaptive algorithm unit 22. The adaptive algorithm unit 22
changes the filter coefficient of the filter unit 21 to change the filter characteristic, thereby
subtracting the output y obtained by filtering the input x from the main input s + na. Adaptive
control to minimize the power of
[0014]
In the specific configuration example shown in FIG. 2, a so-called FIR (finite impulse response)
filter is used as the filter unit 21. In FIG. 2, the reference input x from the input terminal 16a is
sent to a series circuit of delay elements 231, 232,..., 23L according to the number of taps. The
input x0 from the input terminal 16a and the outputs x1, x2, ..., xL from the delay elements 231,
232, ..., 23L are coefficient multipliers 240, 241, 242, ..., 24L, respectively. , And are multiplied
by the filter coefficients w 0, w 1, w 2,..., W L and sent to the adder 25. The filter coefficients w0,
w1, w2,..., WL are corrected by the coefficient correction signal from the adaptive algorithm unit
22, and the output y from the adder 25 is taken out from the output terminal 16b.
[0015]
As the adaptive algorithm used in the adaptive algorithm unit 22, many techniques have been
proposed, and an LMS (least mean square, least mean square) algorithm will be described as a
specific example thereof.
[0016]
The input data at the k-th sample cycle time point (time k) of the data series of the input x and
the delayed output data from each of the delay elements 231, 232, ..., 23L are xk0, xk1, xk2, ... ..,
XkL, an input vector Xk subjected to FIR filter processing is set as Xk = [xk0 xk1 xk2 ... xkL] T (1).
10-04-2019
5
T in this equation (1) represents a transpose. Assuming that the above filter coefficients
(weighting coefficients) are wk0, wk1, wk2,..., WkL for this input vector Xk, and the FIR filter
output is yk, the relationship between input and output is given by the following equation (2) It
will be. yk = wk0xk0 + wk1xk1 +... + wkLxkL (2) Further, if the filter coefficient vector (weighting
vector) Wk is defined as Wk = [wk0 wk1 wk2 wkL] T (3), input / output The relationship is
described as yk = XkTWk (4). Assuming that the desired response is dk, the error .epsilon.k from
the output is expressed as .epsilon.k = dk -yk = dk -Xk T Wk (5). Using these, the LMS algorithm
is expressed as Wk + 1 = Wk + 2μεk Xk (6). (6) is a gain factor that determines the speed and
stability of adaptation.
[0017]
Here, in the embodiment shown in FIG. 1, since the input noise component nb is gain-controlled
by the gain control circuit 19 and sent to the reference input terminal 16a, the reference input x
supplied to the adaptive filter 16 has a gain By the gain g controlled by the control circuit 19, x =
gnb. When the vector of the input noise component corresponding to the input vector Xk
subjected to the FIR filter processing is represented as Nbk, Xk = gNbk, and therefore, the above
equation (6) becomes Wk + 1 = Wk + 2.mu..epsilon.kgNbk (7) It is expressed as Therefore, it can
be understood that by changing the gain g of the gain control circuit 19, the adjustment speed
(the adjustment amount for each sample period) of the filter coefficient vector Wk is changed.
[0018]
As described above, by controlling the level of the reference input x in accordance with the signal
level of the main input detected by the signal level detection circuit 18, the processing speed in
the adaptive filter 16, particularly the correction speed of the filter coefficient Changes.
Specifically, for example, when the main input signal level is large, noise is masked and hardly
heard, so the processing speed is reduced to reduce the noise reduction effect to reduce
distortion, while the main input signal level is reduced. When noise is in the ear when E is small,
etc., it may be considered that the processing speed is increased to actively remove noise. This is
because, if noise reduction processing is performed, noise is not noticeable when the main input
signal level is large and noise can not be heard so much, considering that some distortion occurs
compared to when no processing is performed. While the reduction process is suppressed and
the main input signal is taken out almost as it is, or the filter coefficient modification speed is
slowed to perform only high-precision (less distortion) processing, while noise is offensive to
some extent, It is intended to put more effort into removing noise even if distortion is allowed.
10-04-2019
6
[0019]
Here, FIG. 3 and FIG. 4 show the case of the embodiment of FIG. 1 (FIG. 3) using the
configuration in which the level of the reference input is changed according to such main input
signal level (FIG. 3) and the case of prior art (FIG. 4) is a characteristic diagram showing
respective frequency characteristics. That is, the configuration corresponding to the conventional
example of FIG. 4 is only the adaptive noise reduction circuit configuration from the input
terminals 12 and 16a to the output terminal 17 of FIG. In FIGS. 3 and 4, a sine wave signal of
500 Hz and amplitude 1 is used as the signal component s, and a sine wave signal of amplitude 1
and 1600 Hz is used as the noise components na and nb.
[0020]
As apparent from a comparison of FIGS. 3 and 4, the noise level at about 3 to 4 kHz is reduced by
about 20 dB in the embodiment of FIG. 3 as compared with the prior art of FIG.
[0021]
Next, FIGS. 5 to 7 are graphs for explaining the relationship of the control gain in the gain control
circuit 19 according to the main input signal level detected by the signal level detection circuit
18, and the horizontal axis represents the main As the input signal level, the vertical axis
represents the signal level of the reference input x obtained by gain control by the gain control
circuit 19.
The relationship between the detected main input signal level and the control gain may be
various, such as monotonous increase, monotonous decrease, line graph, etc., and is not limited
to those shown in FIGS. .
[0022]
Here, the example of FIG. 5 shows a case where the main input signal level and the controlled
gain are in a monotonically increasing relationship, and the gain is set such that the reference
input signal level decreases as the main input signal level decreases. Control is made. On the
10-04-2019
7
other hand, in the example of FIG. 6, the main input signal level and the controlled gain are in a
monotonically decreasing relationship, and gain control is performed so as to reduce the
reference input signal level as the main input signal level increases. Ru. At this time, the
processing speed of the adaptive filter 16 is reduced to reduce distortion. Further, in the example
of FIG. 7, the main input signal level is monotonously decreased from 0 to a predetermined
threshold level Th as in the case of FIG. 6 above, and the reference input signal level is set to 0
when the threshold level Th is exceeded. It is designed not to filter. If adaptive filtering is not
performed, the noise in the main input will not be reduced, but if the main input signal level is
large and the noise is masked, it will not be a problem in hearing, but rather by the noise
reduction processing. There is an advantage that distortion of the signal does not occur.
[0023]
The arrangement position of the gain control circuit 19 is not limited to the position shown in
FIG. 1 as long as it can control the speed of the filtering process in the adaptive filter 16. For
example, as shown in FIG. The residual from the adder 15 as means may be inserted into the path
back to the adaptive filter 16. At this time, the error ε k in the equation (6) is subjected to gain
control, and in this case as well, by changing the gain of the gain control circuit 19, the
adjustment speed of the filter coefficient vector Wk It can be seen that the amount) changes.
Therefore, the same effect as the embodiment of FIG. 1 described above can be obtained. The
other parts in FIG. 8 may be configured in the same manner as in FIG. 1 above, and the
corresponding parts will be assigned the same reference numerals and descriptions thereof will
be omitted.
[0024]
The present invention is not limited to the above-described embodiment. For example, the
relationship of the gain controlled with respect to the main input signal level is not limited to the
specific examples of FIGS. Various other things are conceivable. Also, the specific configuration of
the filter unit 21 and the algorithm used for the adaptive algorithm unit 22 are not limited to the
FIR filter and the LMS algorithm of the above embodiment.
[0025]
As is apparent from the above description, according to the noise reduction apparatus of the
10-04-2019
8
present invention, noise reduction processing is adaptively performed using an adaptive filter for
the main input in which noise is mixed in the signal component. Since the processing speed of
the adaptive filter is controlled by detecting the signal level of the main input using a noise
reduction device that performs The degree of adaptive noise reduction processing can be
increased or decreased. Specifically, the noise reduction can be increased when the noise is in the
ear, and the noise reduction can be reduced when the signal level is large and the noise can not
be in the ear, thereby suppressing the generation of distortion due to the processing.
10-04-2019
9
Документ
Категория
Без категории
Просмотров
0
Размер файла
19 Кб
Теги
description, jph0522788
1/--страниц
Пожаловаться на содержимое документа