JP2001344000A - Noise canceler, communication equipment provided with it, and storage medium with noise cancellation processing program stored - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the cancellation performance by excluding an influence of a noise energy to perform stable noise suppression processing even in the case of variance of the noise energy and preventing an unexpected noise from being erroneously detected as a sound to improve the discrimination precision of a sound/noise section. SOLUTION: A minimum value of the noise power in each band is obtained in a noise minimum value estimating circuit 29, and a spectrum shape of this noise minimum value is used for determination of a gain in each band by a band-classified gain determination part 33, and a weighting addition value of a difference in each of plural continuous bands where a threshold exceeds one is obtained by a significant value calculation part 27, and this weighting addition value is used to calculate a sound weight in a sound weight calculation part 28, namely, to discriminate whether a pertinent frame is a sound frame, a noise frame or an intermediate transitional area frame between them.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a noise canceller provided in a communication device for encoding and transmitting a voice signal, such as a digital automobile / portable telephone device, a digital cordless telephone device, a digital wired telephone device, and the like. Related to a communication device.

[0002]

2. Description of the Related Art In a digital portable telephone device, C is generally used.
A low bit rate audio encoding method such as an ELP (Code Excited Linear Prediction) method is used. By using this type of coding, it is possible to make a good voice call even in an environment where background noise is relatively large.
For details of the CELP method, refer to MRSchroeder
And BSAtal's “Code-Excited Linear Prediction
(CELP): High-QualitySpeech At Very Low Bit Rate
s ”in Proc. ICASSP, 1985. pp. 937-939.

However, in a high-noise environment such as a railway platform or a highway, background noise significantly reduces the intelligibility of speech. For this reason, various studies have been made on a noise canceller that removes noise and uses only speech for encoding. One example is the "Enhanced Variable Rate Codec, S" which is a variable rate speech coding standardized in the United States.
peech Service Option 3 for Wideband Spread Spectru
m Digital Systems ”(TIA IS127) has a noise canceller specified as an option.

FIG. 9 is a functional block diagram of a noise canceller defined in the TIA IS127. In the figure, reference numeral 51 denotes a fast Fourier transform unit (FFT).
r transform), and a framed transmission signal is input to the FFT 51. The framed transmission signal is obtained by dividing the A / D-converted transmission signal into frames of, for example, 80 samples, and then dividing the transmission signal by 1 including the overlap.
Generated by shaping into 28 samples. FFT5
1 performs a fast Fourier transform process on the framed transmission signal, thereby obtaining a transform coefficient frequency-divided into, for example, 16 bands.

[0005] The transform coefficients of 16 bands obtained by the FFT 51 are input to a band energy estimating section 52. The band energy estimating unit 52 calculates the energy of the conversion coefficient of the 16 bands, and calculates the calculated value of the energy per band as the band SN.
The signals are input to the R estimation unit 53 and the noise estimation unit 54. The noise estimating unit 54 estimates the energy per band of the noise portion based on the estimated value of the energy per band, and supplies the estimated band-specific noise energy to the band SNR estimating unit 53. Band SNR
The estimation unit 53 calculates a logarithmic value (SNR) of the signal energy and the noise energy for each band based on the calculated value of the energy for each band and the estimated value of the noise energy for each band.
Then, the SNR for each band is calculated by the voice metric calculation unit 5.
Give 5 The voice metric calculation unit 55 calculates the total sum after multiplying the above-mentioned SNR for each band by a weight coefficient corresponding to the magnitude, and the calculation result is a voice metric.

[0006] When the voice metric is smaller than the threshold value, the band SNR correction unit 56 determines that no voice is included, corrects the band SNR to a small value, and provides the band gain calculation unit 57 with the band SNR. The band gain calculator 57 determines the amount of noise suppression for each band based on the corrected SNR,
This is given to the multiplication unit 61. As a result, in the multiplication unit 61, the F
The transform coefficients of the 16 bands output from the FT 51 are multiplied by the above-described noise suppression amount for each band, thereby outputting a signal whose noise is suppressed for each band. The transform coefficients for which noise has been suppressed for each band are inverse fast Fourier transformed by an inverse fast Fourier transform unit (IFFT) 58,
As a result, the signal is returned to the time domain signal and output.

The updating of the noise energy is performed after the spectrum deviation estimating section 59 and the noise updating determining section 60 determine whether or not updating is possible.

That is, in the noise canceller,
The noise suppression amount for each band is determined based on the SNR of the transmission signal, that is, the ratio between the signal energy and the noise energy.
Further, in order to determine whether or not voice is included, a voice metric which is a value obtained by weighting and adding the SNR for each band is used.

[0009]

By the way, noise (Back
ground noise) is generally assumed to be steady, but may fluctuate outdoors. In particular, the energy of noise generated when a vehicle passes by increases as the vehicle approaches. If the transmitted voice is input in this state, the suppressed voice may be distorted because the energy difference between the voice and the noise is small. Also, when the spectral shape of the noise is similar to the spectral shape of the voice, if the suppression is performed based on the noise energy, it becomes easy to interfere with the voice spectrum.
Distortion occurs in the speech after suppression.

The voice metric used for voice detection is
Since the sum is obtained by multiplying the SNR by a weighting factor, if sudden noise is mixed in, it is likely to be regarded as speech.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and a first object of the present invention is to eliminate the influence of noise energy even when the noise energy fluctuates, thereby enabling stable noise suppression processing. An object of the present invention is to provide a noise canceller in which generation of voice distortion due to suppression is reduced, a communication device including the noise canceller, and a storage medium storing a noise cancellation processing program.

A second object of the present invention is to prevent a sudden noise from being erroneously detected as voice, thereby improving the accuracy of determining a voice / noise section and suppressing noise more effectively. Is to provide a storage medium in which is stored.

[0013]

In order to achieve the first object, a first invention divides an input signal into frames of a fixed time length, and divides the signals of these frames into a plurality of frequency bands. A noise canceller that performs noise suppression processing for each of these frequency bands, obtains signal power for each of the frequency bands, and estimates noise power for each band based on the band power; For at least one of the band power and the band-specific noise power, the minimum value detecting means for detecting the minimum value of the power over a plurality of frame periods, and the band power and the minimum value detecting means are detected for each of the frequency bands. And a suppression amount determining means for determining a difference from the minimum value for each band and determining a noise suppression amount for each frequency band based on the difference. Than it is.

[0014] Specifically, the suppression amount determining means may include:
A noise suppression amount determination process for each band is performed based on a difference between the band power and the minimum value for each band.

Therefore, according to the first aspect, the minimum value of the band power or the noise power for each band is detected for each frequency band, and the noise suppression for each frequency band is performed based on the difference between the band power and this minimum value. The amount is determined. Therefore, even if the noise energy temporarily increases during a call, for example, when a car passes, the noise suppression amount is determined based on the difference between the band power and the minimum value, and the suppression amount is stable. , And as a result, stable noise suppression processing can always be performed.

Further, even when the spectrum shape of the noise is similar to the spectrum shape of the voice, the suppression processing is performed based on the difference between the band power and the minimum value. As a result, interference with the spectrum of the voice is reduced, and the occurrence of voice distortion is prevented.

Further, in the present invention, the suppression amount determination means generates an adjustment value common to different bands for each frame, and for each of the frequency bands, the band power and the minimum value for each band detected by the minimum value detection means. It is also characterized in that a difference from a value obtained by adding the adjustment value common to the band to the value is obtained, and the noise suppression amount for each frequency band is determined based on the difference.

Specifically, in a noise section, an adjustment value common to the bands is determined based on a difference between the average value between the minimum values for the respective bands and the average value between the noise powers for the respective bands. Determines an adjustment value common to the bands based on a difference between a minimum value among a plurality of band powers in one frame and a maximum value among a plurality of band-specific minimum values. By doing so, it is possible to determine a more appropriate amount of noise suppression for each of the voice section and the noise section.

On the other hand, in order to achieve the second object, a second invention provides a noise power estimating means for obtaining a signal power for each frequency band and estimating a noise power for each band based on the band power. And comparing means for obtaining a difference between the band power for each frequency band and the noise power for each band estimated by the noise power estimating means, and comparing the difference for each band with a predetermined threshold value. By means, only when it is determined that the band-based difference of a plurality of adjacent bands among the band-based differences arranged in the frequency order exceeds the threshold value, these band-based differences are weighted in a predetermined manner. And a determining means for determining whether the input signal is a voice section or a noise section based on the added value of the band-wise differences obtained by the adding section. .

Specifically, the adding means weights the difference for each band such that the weight decreases as the frequency increases.

The determining means determines whether the input signal is a voice section, a noise section, or a transient section which is an intermediate area between the two sections, based on the added value of the difference for each band obtained by the adding section. It is also possible.

Therefore, according to the second aspect, the speech / noise determination is performed in consideration of the continuity of the band, so that sudden noise is not erroneously detected as speech, thereby providing accurate speech / noise. It is possible to make a determination.

[0023]

FIG. 1 is a circuit block diagram showing an embodiment of a digital portable telephone device provided with a noise canceller according to the present invention.

In FIG. 1, a radio carrier signal transmitted from a base station (not shown) via a radio channel is received by an antenna 1 and then input to a receiving circuit (RX) 3 via an antenna duplexer (DUP) 2. Then, the signal is mixed with the reception local oscillation signal output from the frequency synthesizer (SYN) 4 and frequency-converted into an intermediate frequency signal. The received intermediate frequency signal is sampled by an A / D converter (not shown) and then sampled by a digital demodulator (D
EM) 6.

The digital demodulator 6 performs digital demodulation processing after establishing frame synchronization and bit synchronization with respect to the digital reception intermediate frequency signal. The baseband digital demodulated signal obtained by this demodulation processing is input to a time division multiple access circuit (TDMA) 8, where a time slot addressed to itself is separated and extracted for each transmission frame. The digital demodulator 6
The information relating to the frame synchronization and the bit synchronization obtained in is input to the control circuit 18.

The digital demodulated signal output from the TDMA circuit 8 is subsequently sent to an error correction code decoding circuit (CH-
COD) 9 and is subjected to error correction decoding processing. The digital demodulated signal subjected to the error correction decoding is input to a speech decoding circuit (DEC) 10 and subjected to speech decoding processing, whereby a digital reception signal is reproduced. This digital reception signal is returned to an analog reception signal by the D / A converter 11 and then supplied to a speaker 12 via an audio amplifier (not shown), and the speaker 12 outputs the loudspeaker.

On the other hand, the transmitted voice of the speaker is the microphone 1
The sound is collected at 3 and converted into an electric signal, and then input to the A / D converter 14, where the signal is sampled at a predetermined sampling cycle and converted into a digital transmission signal. The digital transmission signal is passed through a noise canceller 17 to be described later, and then a voice encoding circuit (CO
D) It is input to 16 and voice-encoded.

The coded voice data output from the voice coding circuit 16 is input to an error correction code decoding circuit (CH-COD) 9 together with the control signal output from the control circuit 18, where the error correction coding is performed. Is done. Then, the digital transmission signal subjected to the error correction coding is input to the TDMA circuit 8. In the TDMA circuit 8, a transmission frame corresponding to the time division multiple access (TDMA) system is generated, and processing for inserting the digital transmission signal into a time slot allocated to the own device in the transmission frame is performed. . The digital transmission signal output from the TDMA circuit 8 is subsequently transmitted to a digital modulator (M
OD) 7.

In the digital modulator 7, a transmission intermediate frequency signal digitally modulated by the digital transmission signal is generated, and the transmission intermediate frequency signal is supplied to a D / A (not shown).
After being converted into an analog signal by the converter, it is input to the transmission circuit (TX) 5. As the digital modulation method, for example, π / 4 shifted QPSK (π / 4 shifted
A quadrature phase shift keying method is used.

In the transmission circuit 5, the modulated transmission intermediate frequency signal is first mixed with the transmission local oscillation signal output from the frequency synthesizer 4 and converted into a radio carrier frequency corresponding to the communication channel. The radio modulated wave signal is controlled by the transmission power amplifier to a predetermined transmission power level specified by the control signal TCS from the control circuit 18 and then transmitted from the antenna 1 to the base station (not shown) via the antenna duplexer 2. Sent to.

Reference numeral 19 denotes an operation panel unit. The operation panel unit 19 includes a key input unit having a transmission key, an end key, a dial key, and various function keys, a liquid crystal display (LCD) and a light emitting diode (LED). LED).

The noise canceller 17 is realized by, for example, a DSP (Digital Signal Processor), and its processing program is stored in a memory in the noise canceller or a memory attached to the control circuit 18. FIG. 2 is a block diagram showing a configuration of a function realized by the processing program.

The digital transmission signal output from the A / D converter 14 is first input to the frame division unit 21.
The frame dividing unit 21 divides the digital transmission signal into frames of, for example, 80 samples, and performs windowing to overlap the frame ends,
As a result, a frame adjusted to 128 samples including the overlap is output. Then, the digital transmission signal frame is converted into a fast Fourier transform unit (FFT) 22.
To enter.

The FFT 22 performs a fast Fourier transform process on the input digital transmission signal frame, thereby obtaining, for example, transform coefficients frequency-divided into 16 bands in order from low to high. Note that the number of transform coefficients for each band may not be the same. Then, the band-divided transform coefficients are input to a multiplier 23 for noise suppression processing and to a band power calculator 26.

The band power calculator 26 calculates the energy (mean square value of the transformation variable) for each band and calculates the logarithm,
Output band power channel_power (m, k). Where m
Is a frame number, and k is a band number (k = 0,..., 15).
Then, the band power is input to a noise leak integrated value updating unit 32 and a band-specific gain determining unit 33, which will be described later, and is also input to a significant value calculating unit 27.

The significant value calculation unit 27 includes, for each band, a noise leak integrated value noise_power (m, k) output from a noise leak integrated value updating unit 32 described later, and the band power channel
The difference tmp from l_power (m, k) is calculated, and the difference tm
Compare p to a predetermined threshold. When it is determined that the difference tmp between bands of a plurality of adjacent bands among the difference tmp for each band arranged in the frequency order exceeds the threshold, a predetermined weight is assigned to the difference tmp for each band. And then add each other. Then, the weighted value suby
(m, k) and the conditional sum of this suby are output as a significant value y.

FIG. 3 is a flowchart showing the processing procedure of the significant value calculating section 27 and the contents thereof. In the figure, the significant value calculation unit 27 first resets the frame number m to 0 in step 3a, then increments the group number m in step 3b, and exceeds the significant value y, the band number k, and the threshold value. Consecutive number of difference tmp by band fl
ag is initially set to 0.

Next, in step 3c, for the band k = 0, the significant value calculation unit 27 assigns a weight suby (m) to the difference tmp between the band power and the noise leakage integral value and the difference tmp for each band. , k) are calculated as follows. tmp = channel_power (m, k) -noise_power (m, k) suby (m, k) = {200- (k-1) ² } / 100 * (tmp-1) where {200- (k-1) ² } is a weighting factor.

Then, the band difference tm at the band k = 0.
When p is calculated, the significant value calculation unit 27 compares the band-based difference tmp with the threshold value “1” in step 3d, and determines the band-based difference.
If tmp exceeds the threshold value “1”, it is determined that there is a possibility that the sound is voice, and the process proceeds to step 3i via steps 3e and 3g, where the continuous number flag = 1 is set. Then, after incrementing the band number k to k = 1 in step 3k, the process returns to step 3c, and the same processing is executed for the band k = 1.

Now, even in this band k = 1, the band
It is assumed that the band difference tmp exceeds the threshold “1” following k = 0. Then, the significance value calculation unit 27 moves from step 3e to step 3f because flag = 1 has already been set, and performs an operation of y = y + suby (m, k-1). That is, the suby in the band k = 0
Let (m, k-1) be a significant value y. Then, after setting the continuous number flag = 2, the process proceeds to step 3h via step 3g,
Here, an operation of y = y + suby (m, k) is performed, whereby the suby in the band k = 0 is obtained.
The suby (m, k) obtained in the current band k = 1 is added to (m, k-1). Then, in step 3k, the band number k is further incremented to be k = 2, and the process returns to step 3c to execute the processing for band k = 2.

Thereafter, similarly, adjacent bands k = 2, k = 3, k = 4
,... Each time the difference tmp of each band exceeds the threshold value “1”, the suby (m, k) of the band is sequentially added to the significant value y obtained up to the immediately preceding band, whereby the band A weighted addition value y of the difference tmp is obtained.

When the difference tmp for each band becomes equal to or less than the threshold value "1" in any band k = i, the significant value calculating section 27 shifts from step 3d to step 3j, where
Reset flag to 0.

When the processing for all 16 bands k = 0 to k = 15 constituting one frame is completed, the significant value calculation unit 27 shifts from step 3m to step 3n, where The significant value y and the weighted difference suby (m, k) (k = 0,1,..., 15) calculated for each band are output.

Thus, for each frame, a weighted addition value y of the band-by-band difference tmp of a plurality of continuous bands whose threshold value exceeds "1" is obtained, and this weighted addition value y is used as a voice weight calculation unit to be described later. The speech weight is calculated at 28, that is, whether the frame is a speech frame or a noise frame, and whether the frame is an intermediate transition frame. That is, when the band-based difference tmp exceeds the threshold value “1” in only one band, this is regarded as noise and eliminated, and the voice / noise / transient region in the voice weight calculation unit 28 is determined. It is not used for judgment.

The weighted addition value from the significant value calculation unit 27
When y is supplied, the speech weight calculator 28 calculates the speech weight sp used for determining the noise suppression gain. The audio weight sp is 0 ≦ the degree of audio included in one frame.
a numerical value expressed in a range of sp ≦ 6, wherein the weighted addition value y
Is calculated from Note that sp = 0 represents a noise section, and sp = 6 represents a speech section.

FIG. 4 is a flowchart showing the procedure for calculating the speech weight sp in the speech weight calculator 28 and the processing contents thereof. First, the voice weight calculation unit 28 performs step 4a.
After resetting the frame number m to 0 in step 4,
The group number m is incremented by b. Next, in step 4c, the weighted addition value y is set to an arbitrary threshold value “1”.
3 and if y <13, it is determined that the frame is a noise frame, and the flow shifts to step 4d where the speech weight sp is set to sp (m) = sp (m-1) -0.5. On the other hand, if y ≧ 13, the process proceeds to step 4e, where z = (y−13) * 1.5 + 1 is calculated. That is, the speech weight z is set to 1 to 6 based on y.
Set temporarily within the range.

Subsequently, the voice weight calculation unit 28 determines in step 4
At f, it is determined that sp (m-1) ≦ 0.5. That is, it is determined whether or not the speech weight sp (m-1) one frame before is a noise frame. If the frame is a noise frame, the process proceeds to step 4g, where the speech weight sp of the current frame is entered.
(m) is set to the above z. On the other hand, if the speech weight sp (m-1) one frame before is not a noise frame, the process proceeds to step 4h, where z> sp (m-1) +0.5 is determined, and z If> sp (m-1) +0.5, the voice weight sp (m) of the current frame is set to sp (m-1) +0.5 in step 4i. On the other hand, if it is not z> sp (m-1) +0.5, the process proceeds to step 4j, where z> sp (m-1) -0.5 is determined, and z> sp (m-1) -0.5
If 0.5, the voice weight sp (m) of the current frame in step 4k
Is set to sp (m-1) -0.5. Also, z> sp (m-1) −0.5
If not, the process proceeds to step 4m, where the speech weight of the current frame is set to sp (m) = MIN (sp (m), 6) or sp (m) =
Set to MAX (sp (m), 0).

That is, in steps 4f to 4m, the temporary voice weight z calculated in the current frame is used.
Is corrected in consideration of the voice weight sp (m-1) set in the immediately preceding frame. Therefore, by using the speech weight sp (m) obtained in this way, it is possible to determine the speech / noise / transient region in consideration of the continuity between frames.

The speech weight sp (m) obtained by the speech weight calculator 28 is output in step 4n, and is input to the noise minimum value estimator 29 and the band-specific gain determiner 33.

The noise minimum value estimating unit 29 determines that the speech weight is
The minimum value of the noise integrated value noise_power (m, k) in each band is checked every 100 frames in which sp = 0. Then, this minimum value is used as the noise minimum value noise_min (m, k) in the period of the next 100 frames.
At the same time, the inter-band average value of the minimum noise value of each band, mi
Find n_all.

FIGS. 5 and 6 show the noise minimum value estimating unit 2.
9 is a flowchart showing the procedure and contents of a minimum value estimation process executed in FIG. In the figure, the noise minimum value estimating section 29 firstly sets the frame number m to m = m in step 5a.
And reset the frame counter value to fc
= 96, the noise minimum to noise_min (k) = 36, and the bandwidth
Initially set to k = 0, ..., 15, and noise_min
(k) _h (k) = MAX (noise_power (m, 2k), noise_power (m, 2k +
1)), k = 0, ..., 7, the average of the minimum noise values between bands min_all Initialize each.

Next, the noise minimum value estimating unit 29 determines in step 5
After incrementing the frame number m in b, it is determined in step 5c whether the speech weight is sp = 0, that is, whether the frame is a noise frame. If the frame is a noise frame, the flow returns to step 5b to increment the frame number m, and the noise frame is determined in step 5c. That is, a voice frame or a transition area frame is detected in steps 5b and 5c.

When a voice frame or a transient frame is detected, the minimum noise value estimating unit 29 proceeds to step 5d, where the frame counter fc is incremented and a band k = 0 is selected. And step 5
After e is set to x = MAX (noise_power (m, 2k), noise_power (m, 2k + 1)), the process proceeds to step 5f and noise_min (k)
_h (k)> x is determined, and noise_min (k) _h (k)> x
If so, the process proceeds to step 5g, where the minimum noise value is set to no.
Set ise_min (k) _h (k) = x. And step 5
h.

On the other hand, if noise_min (k) _h (k)> x is not satisfied, the process directly proceeds to step 5h to select the next band k = 1, and the above steps 5e to 5e are performed until the band k = 8 is reached. The setting process of the noise minimum value noise_min (k) _h (k) by 5g is repeated.

When the band reaches k = 8, the minimum noise value estimating unit 29 sets the frame counter fc to 1 at step 5j.
It is determined whether 00 has been reached. Until the number of frames reaches 100, the process returns to step 5b to select the next frame.
The processing from c to step 5i is repeated.

On the other hand, when the processing for the above 100 frames is completed, the noise minimum value estimating section 29 proceeds to step 6a shown in FIG. 6, where the average noise minimum value min_al
l Ask by

Also, noise_min (0) and noise_min
noise_min (0) = noise_min_h (0) noise_min (1) = 0.75 noise_min_h (0) + 0.25 noise_min_
h (1) and the band is k = 1.

Further, the noise minimum value estimating section 29 proceeds to step 6b, where the remaining band k = k based on the eight noise minimum values previously obtained for the bands k = 0 to k = 7. Noise_min (2k) = 0.75 noise_min_h (k) + 0.25 noise_min
_h (k-1) noise_min (2k + 1) = 0.75 noise_min_h (k) +0.25 noise_m
Calculate as in_h (k + 1).

When the above operation is completed, the noise minimum value estimating section 29 proceeds from step 6d to step 6e, where noise_min (14) = 0.75 noise_min_h (7) +0.25 noise_min
_h (6) Calculate noise_min (15) = noise_min_h (7).

That is, the minimum noise value estimating section 29 interpolates 16 min_alls based on the eight min_alls in the above-mentioned steps 6a to 6e.

After calculating the 16 min_all values, the minimum noise value estimating unit 29 resets the frame counter fc to 0 and sets the minimum noise value in step 6f.
Reset noise_min_h (k) = 36 and band to k = 0, ..., 7. Then, in step 6g, the inter-band average value min_all of the previously calculated noise minimum value and the noise minimum value noise_m
After outputting in (m, k), k = 0,..., 15, the process returns to step 5b to repeat the same process of calculating the minimum noise value and the inter-band average value for the next frame m = m + 1.

The update determining unit 31 and the noise leak integrated value updating unit 32 perform a noise leak integrated value noise_power.
The updating process of (m, k) is performed. That is, the update determination unit 3
1 indicates that updating is possible when y <15, and updating is not possible otherwise. Noise leak integration value update unit 3 when update is possible
2 is the noise power noise_power, for example, noise_power (m + 1, k) = noise_power (m, k) * 0.9 + channel_
Update as power (m, k) * 0.1, k = 0, ..., 15.

The band-specific gain determining section 33 outputs the band power channel_power output from the band power calculating section 26.
(m, k), the noise power noise_power (m, k) output from the noise leak integration value update unit 32, the voice weight sp (m, k) output from the voice weight calculation unit 28, and the noise minimum value estimation unit 29
Based on the noise minimum value noise_min (m, k) output from, the gain for each band gain (m, k) is determined.

First, the band average value noise_all of the noise leak integrated value noise_power (m, k) is calculated as follows: Ask by

Subsequently, the band minimum value min_band of the band power channel_power (m, k) and the band maximum value max_band of the noise minimum value noise_min (m, k) are respectively defined as min_band = MIN (channel_power (m, k), k) = 2, ..., 11 max_band = MAX (noise_power (m, k), k = 0, ..., 15)

Next, the adjustment value md common to the bands is calculated as md = (noise_all−min_all) * (1−sp / 6) + (min_band−max_
band) * sp / 6. According to this equation, md = noise_all−min_all sp = 6, ie, md = min_band−max_band, when sp = 0, that is, in a noise section, and it is understood that the transition region takes an intermediate value between these.

FIGS. 7 and 8 show examples of frequency versus power characteristics for a noise frame and a speech frame, respectively. In a noise frame, as shown in FIG. 7, the band power approaches the noise minimum. The value obtained by adding the adjustment value to the noise minimum value is a value obtained by changing the average value to the noise noise average value noise_all while maintaining the spectral characteristics of the noise minimum value.

On the other hand, in the case of a voice frame, FIG.
As shown in (2), the value obtained by adding the adjustment value to the noise minimum value is adjusted so that the maximum value of the band coincides with the minimum value of the band power while maintaining the spectral characteristics of the minimum value.

The gain per band gain (m, k) is the band power cha
It is determined as follows from nnel_power (m, k), the minimum noise value noise_min (m, k), and the adjustment value. That is, in band k, tmp = channel_power (m, k) −noise_min (m, k) −md−1.625 gain (m, k) = {sqrt (1.4 + 0.49 * tmp2) + 0.7 * tmp−9.5} * 2 is obtained independently for k = 0,. Note that Sp =
When 0 is noise, the gain may be set to a constant in all bands, for example, gain (m, k) =-20, regardless of the above formula.

The band-specific gain gain (m, k) obtained as described above is multiplied by a transform coefficient for each band in the multiplier 23, thereby performing noise cancellation. Then, the transform coefficients for each band, for which the noise has been canceled, are subjected to inverse fast Fourier transform in IFFT 24 and returned to signal frames on the time axis, and are then frame-synthesized in frame synthesizing section 25 to speech encoding circuit 16. Supplied.

As described above, according to this embodiment,
The noise minimum value estimating circuit 29 obtains the minimum value of the noise power of each band, and the spectrum shape of the noise minimum value is used for the determination of the band-specific gain by the band-specific gain determination unit 33. It is possible to realize a noise canceling process that is not affected by a short-term change in the noise spectrum as in the case and is unlikely to distort the voice spectrum.

Further, according to this embodiment, the significance value calculation unit 27 obtains a weighted addition value of the band-by-band difference of a plurality of continuous bands whose threshold value exceeds “1”, and uses this weighted addition value as the voice weight. The calculation of the voice weight in the calculation unit 28, that is, the determination of whether the frame is a voice frame or a noise frame, and further, whether the frame is an intermediate transition area frame. For this reason, when the band-by-band difference exceeds the threshold “1” in only one band, this can be regarded as noise and eliminated, thereby making it possible to accurately determine the voice / noise / transient region. To improve the noise canceling performance.

The present invention is not limited to the above embodiment. For example, in the above-described embodiment, a digital mobile phone device adopting the TDMA system has been described as an example, but the present invention is also applicable to a digital mobile phone device adopting the CDMA system.

In addition, the processing procedure and processing contents of each functional unit in the noise canceller, and the circuit configuration or processing program for realizing this processing can be variously modified and implemented without departing from the gist of the present invention. .

[0075]

As described above, according to the first aspect, the minimum value of the band power or the noise power for each band is detected for each frequency band, and the difference between the band power and this minimum value is determined. By determining the amount of noise suppression for each frequency band, even if the noise energy temporarily increases due to the passage of a car or the like during a call, the amount of suppression can be stably maintained. It is possible to provide a noise canceller capable of reducing the occurrence of voice distortion due to suppression, a communication device including the noise canceller, and a storage medium storing a noise cancellation processing program.

According to the second invention, the difference between the band power for each frequency band and the noise power for each band estimated by the noise power estimating means is determined, and the difference for each band is determined by a predetermined threshold. Only when it is determined that the band-by-band difference between a plurality of adjacent bands among the band-by-band differences arranged in the frequency order exceeds the threshold value, these band-by-band differences are weighted in a predetermined manner. Then, the sum is obtained, and it is determined whether the speech section or the noise section is based on the weighted addition value of the band difference, so that sudden noise can be prevented from being erroneously detected as speech. It is possible to provide a storage medium that stores a noise cancellation processing program capable of increasing the determination accuracy of a voice / noise section and more effectively suppressing noise.

[Brief description of the drawings]

FIG. 1 is a circuit block diagram showing an embodiment of a digital mobile phone device provided with a noise canceller according to the present invention.

FIG. 2 is a functional block diagram showing an embodiment of a noise canceller according to the present invention.

FIG. 3 is a flowchart showing a processing procedure and its contents in a significant value calculation unit of the noise canceller shown in FIG. 2;

FIG. 4 is a flowchart showing a processing procedure and its contents in a voice weight calculator of the noise canceller shown in FIG. 2;

FIG. 5 is a flowchart showing a processing procedure in a noise minimum value estimating unit of the noise canceller shown in FIG. 2 and the first half of its contents;

FIG. 6 is a flowchart showing a processing procedure in a noise minimum value estimating unit of the noise canceller shown in FIG. 2 and a latter half of its contents.

FIG. 7 is a diagram illustrating a relationship between band power, a noise minimum value, and an adjusted noise minimum value in a noise section.

FIG. 8 is a diagram showing a relationship between band power, a noise minimum value, and a noise minimum value after adjustment in a voice section.

FIG. 9 is a diagram showing an example of a configuration of a conventional noise canceller.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Antenna 2 ... Antenna duplexer (DUP) 3 ... Receiving circuit (RX) 4 ... Frequency synthesizer (SYN) 5 ... Transmitting circuit (TX) 6 ... Digital demodulator (DEM) 7 ... Digital modulator (MOD) 8 ... Time division multiple access circuit (TDMA) 9 Error correction code decoding circuit (CH-COD) 10 Voice decoding circuit (DEC) 11 D / A converter 12 Speaker 13 Microphone 14 A / D converter 16 Voice coding circuit (COD) 17 Noise canceller 18 Control circuit 19 Operation panel unit 21 Frame division unit 22 Fast Fourier transform unit (FFT) 23 Multiplier unit 24 Inverse fast Fourier transform unit (IFFT) 25 Frame Synthesizing unit 26 Band power calculating unit 27 Significant value calculating unit 28 Speech weight calculating unit 29 Noise minimum value estimating unit 31 Update determining unit 32: Noise leak integrated value updating unit 33: Gain determining unit for each band

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H03H 21/00 G10L 9/00 D H04B 3/20 9/16

Claims (11)

[Claims] 1. A noise canceller that divides an input signal into frames of a fixed time length, divides signals of these frames into a plurality of frequency bands, and performs noise suppression processing for each of the frequency bands. Noise power estimating means for estimating noise power for each band based on the band power, and for at least one of the band power and the noise power for each band, over a plurality of frame periods. Minimum value detecting means for detecting a minimum value of power; and obtaining a difference between the band power and the minimum value for each band detected by the minimum value detecting means for each of the frequency bands. And a suppression amount determining means for determining a noise suppression amount. 2. The apparatus according to claim 1, further comprising: a unit that determines a voice section and a noise section for the input signal, wherein the suppression amount determining unit determines the band power and the noise level in at least a voice section of the voice section and the noise section. The noise canceller according to claim 1, wherein a noise suppression amount determination process for each band is performed based on a difference from a minimum value for each band. 3. The suppression amount determining means further includes an adjustment value generating means for generating an adjustment value common to different bands for each frame, and for each of the frequency bands, the band power thereof;
A difference between the minimum value for each band detected by the minimum value detection means and a value obtained by adding an adjustment value common to the bands generated by the adjustment value generation means is obtained, and noise suppression for each frequency band is determined based on the difference. The noise canceller according to claim 1, wherein the amount is determined. 4. The adjustment value generation means determines an adjustment value common to bands in a noise section based on a difference between an average value between the minimum values for each band and an average value between noise powers for each band. On the other hand, in an audio section, an adjustment value common to bands is determined based on a difference between a minimum value among a plurality of band powers in one frame and a maximum value among a plurality of band-specific minimum values. The noise canceller according to claim 3. 5. A noise canceller that divides an input signal into frames of a fixed time length, divides signals of these frames into a plurality of frequency bands, and performs noise suppression processing for each of these frequency bands. Noise power estimating means for obtaining signal power for each frequency band and estimating noise power for each band based on the band power; and band-specific noise estimated by the noise power estimating means for each frequency band. A comparing means for determining a difference from the power and comparing these band-by-band differences with a predetermined threshold value; An adding means for, when it is determined that the difference exceeds the threshold, adding these weights to each other after performing predetermined weighting, Based on the sum of the obtained per-band differential by, the noise canceller, characterized by comprising a determining means for determining a speech section or a noise section for the input signal. 6. The noise canceller according to claim 5, wherein said adding means weights the difference for each band such that the weight decreases as the frequency increases. 7. The determining means determines whether the input signal is a voice section, a noise section, or a transient section which is an intermediate area between the two sections, based on an addition value of the difference for each band obtained by the adding section. The noise canceller according to claim 5, wherein the determination is performed. 8. A communication apparatus for transmitting a transmission input signal by encoding the transmission input signal by a voice encoding unit, wherein the transmission input signal is divided into frames having a fixed time length, and the signals of these frames are respectively transmitted to a plurality of frequency bands. A noise canceller that performs noise suppression processing for each of these frequency bands, obtains signal power for each of the frequency bands, and estimates noise power for each band based on the band power. Noise power estimating means, at least one of the band power and the band-specific noise power, a minimum value detecting means for detecting a minimum value of power over a plurality of frame periods, and a band power for each of the frequency bands. The difference from the minimum value for each band detected by the minimum value detecting means is obtained, and the noise suppression amount for each frequency band is calculated based on the difference. Communication apparatus characterized by comprising a constant suppressing amount determining means. 9. A communication apparatus for transmitting a transmission input signal by encoding the transmission input signal by a voice encoding unit, wherein the transmission input signal is divided into frames of a fixed time length, and the signals of these frames are respectively transmitted to a plurality of frequency bands. A noise canceller that performs noise suppression processing for each of these frequency bands, obtains signal power for each of the frequency bands, and estimates noise power for each band based on the band power. Noise power estimating means for determining, for each frequency band, a difference between the band power and the noise power for each band estimated by the noise power estimating means, and comparing these band differences with a predetermined threshold value. Means for determining that the difference between the bands of a plurality of adjacent bands exceeds the threshold value among the differences for each band arranged in order of frequency. And a summation means for weighting these band-by-band differences and then adding them to each other, based on the sum of the band-by-band differences obtained by the addition means. A communication device comprising: a determination unit configured to determine whether the signal is a noise section. 10. A noise canceling program for dividing an input signal into frames of a fixed time length, dividing signals of these frames into a plurality of frequency bands, and performing noise suppression processing for each of these frequency bands. Determining the power of the signal for each frequency band, estimating noise power for each band based on the band power; and at least one of the band power and the noise power for each band, Detecting a minimum value of power over a frame period of; and obtaining a difference between the band power and the detected minimum value for each band for each of the frequency bands, and based on the difference, noise suppression for each frequency band. A storage medium storing a noise canceling processing program comprising the step of determining an amount. 11. A noise cancellation processing program for dividing an input signal into frames of a fixed time length, dividing signals of these frames into a plurality of frequency bands, and performing noise suppression processing for each of these frequency bands. Determining the power of the signal for each frequency band, and estimating noise power for each band based on the band power; and for each frequency band, the band power and the estimated noise power for each band. And comparing these band-by-band differences with a predetermined threshold value. The comparison shows that the band-by-band difference of a plurality of adjacent bands among the band-by-band differences arranged in the frequency order is When it is determined that the difference exceeds the threshold value, adding a predetermined weight to these band-by-band differences and adding them to each other; Based on the sum of more resulting per-band differential, the storage medium storing the noise cancellation processing program and a step of determining whether the speech segment or noise segment for the input signal.