CN117437931B - Sound signal optimized transmission method for microphone - Google Patents
Sound signal optimized transmission method for microphone Download PDFInfo
- Publication number
- CN117437931B CN117437931B CN202311754560.1A CN202311754560A CN117437931B CN 117437931 B CN117437931 B CN 117437931B CN 202311754560 A CN202311754560 A CN 202311754560A CN 117437931 B CN117437931 B CN 117437931B
- Authority
- CN
- China
- Prior art keywords
- signal
- imf
- segment
- sound
- sound signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000005236 sound signal Effects 0.000 title claims abstract description 190
- 238000000034 method Methods 0.000 title claims abstract description 70
- 230000005540 biological transmission Effects 0.000 title claims abstract description 22
- 238000002955 isolation Methods 0.000 claims description 44
- 238000004364 calculation method Methods 0.000 claims description 14
- 238000010606 normalization Methods 0.000 claims description 12
- 230000011218 segmentation Effects 0.000 claims description 7
- 230000003044 adaptive effect Effects 0.000 claims description 5
- 238000007476 Maximum Likelihood Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 abstract description 4
- 230000000694 effects Effects 0.000 description 6
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
- G10L21/0216—Noise filtering characterised by the method used for estimating noise
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Quality & Reliability (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Circuit For Audible Band Transducer (AREA)
Abstract
The invention relates to the technical field of voice processing, in particular to a voice signal optimization transmission method for a microphone, which comprises the following steps: collecting a sound signal through a microphone, and acquiring a plurality of IMF components according to the sound signal; acquiring the possibility that a signal segment in each IMF component is a noise signal according to the sound signal and the IMF component; acquiring the possibility that each sound signal segment is simultaneously uttered by multiple people according to the possibility that the signal segment in each IMF component is a noise signal; according to the possibility that each sound signal segment is simultaneously uttered by multiple people, adaptively generating the threshold value of wavelet threshold denoising of each sound signal segment; and denoising the sound signal according to the threshold value of wavelet threshold denoising of each sound signal segment. According to the invention, the sound signals collected by the microphone are decomposed and segmented, so that the possibility that each segment of sound signals are uttered by a person is obtained, and the signal characteristics of the person when the person utters are prevented from being lost when the sound signals are denoised.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a sound signal optimized transmission method for a microphone.
Background
Since the sound signal of the microphone is often affected by the external environment to generate noise, the quality of the sound signal of the microphone is degraded, and in order to improve the quality of the sound signal, the noise signal in the sound signal needs to be removed. The traditional method for denoising the sound signal is wavelet threshold denoising; however, in the actual situation, a situation that multiple people speak simultaneously can be caused, signals with multiple frequencies are mutually overlapped, so that the mode is disordered, and the wavelet threshold denoising with a fixed threshold cannot take the value of the sound signal when the multiple people speak simultaneously, so that a good denoising effect cannot be achieved.
Disclosure of Invention
The invention provides a sound signal optimization transmission method for a microphone, which aims to solve the existing problems: the traditional wavelet threshold denoising method with fixed threshold can not take the value of sound signals when a plurality of people speak at the same time, and has good denoising effect.
The invention relates to a sound signal optimization transmission method for a microphone, which adopts the following technical scheme:
the method comprises the following steps:
collecting a sound signal through a microphone, and decomposing the sound signal to obtain a plurality of IMF components;
segmenting the sound signal and the IMF components to obtain a signal segment and a sound signal segment in each IMF component; obtaining a noise section of each IMF component according to the amplitude of each sound signal section and the time corresponding relation between each sound signal section and the signal section of each IMF component; acquiring the possibility that the signal segment in each IMF component is a noise signal according to the difference between each signal segment in each IMF component and the noise signal segment and the amplitude of each signal segment in each IMF component;
acquiring the isolation degree of the signal segment in each IMF component according to the possibility that the signal segment in each IMF component is a noise signal and the difference between the amplitudes; acquiring the possibility that each sound signal segment is simultaneously uttered by multiple people according to the isolation degree of the signal segment in each IMF component and the possibility that the signal segment in each IMF component is a noise signal;
according to the possibility that each sound signal segment is simultaneously uttered by multiple people, adaptively generating the threshold value of wavelet threshold denoising of each sound signal segment; and denoising the sound signal according to the threshold value of wavelet threshold denoising of each sound signal segment.
Preferably, the method for acquiring the sound signal through the microphone and decomposing the sound signal to obtain a plurality of IMF components includes the following specific steps:
the sound signal is collected by a microphone and then decomposed by using an EMD algorithm to obtain a plurality of IMF components.
Preferably, the segmenting the sound signal and the IMF component, to obtain a signal segment and a sound signal segment in each IMF component, includes the following specific methods:
presetting a segment threshold valueDividing each IMF component signal and sound signal into sections of length +.>And obtaining the signal segment and the sound signal segment in each IMF component, and completing the segmentation of each IMF component and the segmentation of the sound signal.
Preferably, the method for obtaining the noise segment of each IMF component according to the amplitude of each sound signal segment and the time correspondence between each sound signal segment and the signal segment of each IMF component includes the following specific steps:
calculating all amplitude averages in each sound signal section, marking the sound signal section with the smallest amplitude average as a noise signal section, marking the time section corresponding to the noise signal section as a noise time section, and marking the IMF component signal section of each IMF component under the noise time section as the noise section of each IMF component.
Preferably, the obtaining the probability that the signal segment in each IMF component is a noise signal according to the difference between each signal segment in each IMF component and the noise signal segment and the amplitude of each signal segment in each IMF component includes the following specific methods:
for the firstThe (th) of the IMF components>The signal section is first obtained +.>The (th) of the IMF components>Variance and mean of all amplitudes in the signal section, combined with +.>All the amplitude averages in the noise section of the IMF component, obtain +.>The (th) of the IMF components>The probability that each signal segment is a noise signal is specifically calculated by the following formula:
；
in the method, in the process of the invention,indicate->The (th) of the IMF components>The likelihood that the individual signal segments are noise signals; />Indicate->The (th) of the IMF components>Variance of all amplitudes in the individual signal segments; />Indicate->The (th) of the IMF components>The average of all amplitudes in the individual signal segments; />Indicate->All the amplitude averages in the noise section of the IMF components; />Representing a linear normalization function; />Representing an absolute value operation.
Preferably, the method for obtaining the isolation degree of the signal segment in each IMF component according to the probability that the signal segment in each IMF component is a noise signal and the difference between the amplitudes includes the following specific steps:
for calculation of the firstThe (th) of the IMF components>The isolation degree of each signal segment is firstly obtainedThe mean value of all amplitudes in each signal segment, the +.>The probability that the signal segment is a noise signal is based on the +.>The mean value of all amplitudes in each signal segment, the +.>The possibility that the signal segment is a noise signal, obtain +.>The (th) of the IMF components>Isolation of individual signal segments.
Preferably, the acquiring a firstThe (th) of the IMF components>The isolation degree of each signal segment comprises the following specific calculation formulas:
；
in the method, in the process of the invention,indicate->The (th) of the IMF components>Isolation of individual signal segments; />Representing the number of IMF components;indicate->The (th) of the IMF components>The average of all amplitudes in the individual signal segments; />Indicate->The (th) of the IMF components>The average of all amplitudes in the individual signal segments; />Indicate->The (th) of the IMF components>The likelihood that the individual signal segments are noise signals; />Indicate->The (th) of the IMF components>The likelihood that the individual signal segments are noise signals; />Representing a linear normalization function; />Representing an absolute value operation.
Preferably, the method for obtaining the possibility that each sound signal segment is simultaneously uttered by multiple people according to the isolation degree of the signal segment in each IMF component and the possibility that the signal segment in each IMF component is a noise signal includes the following specific steps:
for calculation of the firstThe individual sound signal segments are at the possibility of multiple people speaking simultaneously; according to the first of all IMF componentsIsolation of individual signal segments, th +.>The possibility that the signal segment is a noise signal, obtain +.>The specific calculation formula of the sound signal segments is as follows:
；
in the method, in the process of the invention,indicate->The individual sound signal segments are at the possibility of multiple people speaking simultaneously; />Represents the +.o of all IMF components>The maximum value of the degree of isolation in the individual signal segments; />Represents the +.o of all IMF components>Minimum degree of isolation in individual signal segments; />Represents the +.o of all IMF components>The next highest value of isolation in the individual signal segments;represents the +.o of all IMF components>The maximum likelihood of a noise signal in the signal segments; />Representing a linear normalization function.
Preferably, the adaptive generation of the threshold size of wavelet threshold denoising of each sound signal segment according to the possibility that multiple people are talking at the same time comprises the following specific methods:
for adaptive generation of the firstThe threshold value of wavelet threshold denoising of each sound signal segment is first preset with an initial wavelet threshold +.>The method comprises the steps of carrying out a first treatment on the surface of the Based on +.>The degree of isolation of the individual signal segments, and the likelihood of each sound signal segment being in simultaneous speech by multiple persons, adaptively generates the +.>The specific calculation formula of the wavelet threshold denoising threshold of each sound signal segment is as follows:
；
in the method, in the process of the invention,indicate->A wavelet threshold denoising threshold for each sound signal segment; />Representing the +.f among all IMF components>The maximum value of the degree of isolation in the individual signal segments; />Indicate->The individual sound signal segments are at the possibility of multiple people speaking simultaneously; />Representing a linear normalization function.
Preferably, the denoising method for the sound signal according to the threshold value of the wavelet threshold denoising for each sound signal segment comprises the following specific steps:
and carrying out wavelet threshold denoising processing on each sound signal segment according to the threshold size of wavelet threshold denoising of each sound signal segment, and completing denoising of the sound signal.
The technical scheme of the invention has the beneficial effects that: according to the invention, each IMF component signal segment and each sound signal segment are obtained by decomposing and segmenting the sound signal, and the possibility that each sound signal segment is simultaneously talking by a plurality of people is obtained by analyzing the characteristics of the sound signal when the plurality of people speak, so that the threshold value of wavelet threshold denoising is adaptively generated according to the possibility, the signal characteristics of people speaking are prevented from being lost when the sound signal is denoised, and the purpose of improving the sound signal quality is achieved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flow chart of steps of a method for optimized transmission of sound signals for microphones according to the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following description refers to the specific implementation, structure, characteristics and effects of an optimized transmission method for sound signals of microphones according to the invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the sound signal optimization transmission method for the microphone provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of steps of a method for optimizing transmission of sound signals of a microphone according to an embodiment of the present invention is shown, the method includes the following steps:
step S001: the sound signal is collected through the microphone, and the sound signal is decomposed to obtain a plurality of IMF components.
It should be noted that, since the sound signal of the microphone is often affected by the external environment to generate noise, the quality of the sound signal of the microphone is reduced, but the final objective of the embodiment is to improve the quality of the sound signal as a sound signal optimizing transmission method for the microphone; for this purpose, the sound signal of the microphone needs to be recorded first.
Specifically, the sound signal is collected by a microphone.
It should be further noted that, in order to be able to perform denoising on the sound signal better, it is also necessary to decompose the sound signal, and prepare for denoising the sound signal later.
Specifically, the sound signal is decomposed by using the EMD algorithm to obtain a plurality of IMF components, and since the specific process of the EMD algorithm decomposition is a well-known prior art, a detailed description is omitted in this embodiment.
To this end, several IMF components are obtained.
Step S002: segmenting the sound signal and the IMF components to obtain a signal segment and a sound signal segment in each IMF component; obtaining a noise section of each IMF component according to the amplitude of each sound signal section and the time corresponding relation between each sound signal section and the signal section of each IMF component; and acquiring the possibility that the signal segment in each IMF component is a noise signal according to the difference between each signal segment in each IMF component and the noise signal segment and the amplitude of each signal segment in each IMF component.
It should be noted that, the final objective of this embodiment as a method for optimizing transmission of sound signals for microphones is to improve the quality of sound signals, i.e. to remove noise signals from the sound signals. The traditional method for denoising the sound signals is wavelet threshold denoising, but because the condition that a plurality of people speak simultaneously exists in the actual situation, signals with a plurality of frequencies are mutually overlapped, so that the mode is disordered, and the wavelet threshold denoising with a fixed threshold cannot take the value of the sound signals when the plurality of people speak simultaneously, so that a good denoising effect cannot be achieved. The present embodiment proposes a denoising method for a sound signal.
It should be further noted that, in the denoising method for the sound signal provided in this embodiment, a local range of each IMF component is first obtained, local characteristics of each local range are analyzed, a probability that each local range is noise data is calculated according to the local characteristics of each local range, and an isolation degree of each local range is calculated according to the probability that each local range is noise data; and calculating the possibility that each segment is simultaneously talking for multiple people according to the isolation degree of each local range, and adaptively generating the threshold value of wavelet threshold denoising. Since the sound signal has stable characteristics for a short period of time, it is necessary to segment each IMF component as well as the sound signal in order to better capture the local characteristics of the sound signal.
Specifically, a segmentation threshold value is preset，/>The specific size of (2) can be set by the user according to the actual situation, the hard requirement is not required in the embodiment, and +.>Millisecond to describe; dividing each IMF component signal and sound signal into sections of length +.>And obtaining the signal segment and the sound signal segment in each IMF component, and completing the segmentation of each IMF component and the segmentation of the sound signal.
It should be noted that, in the sound signal, there are a signal when a person speaks and a signal when the person does not speak, that is, the signal when the person speaks in the sound signal is not uniformly distributed in the sound signal, but the noise signal is uniformly distributed in the sound signal; therefore, the signal when not speaking is a noise signal; and because the amplitude of the signal when the person speaks is larger than the amplitude of the noise signal; so that the noise signal can be obtained based on this.
Specifically, all the average amplitude values in each sound signal segment are calculated, the sound signal segment with the smallest average amplitude value is recorded as a noise signal segment, the time segment corresponding to the noise signal segment is recorded as a noise time segment, and the IMF component signal segment of each IMF component under the noise time segment is recorded as the noise segment of each IMF component.
It should be noted that, the amplitude of the signal when a person speaks in the sound signal is greatly changed, and the amplitude of the noise signal is always stable, so the probability that the signal segment in each IMF component is the noise signal can be calculated according to the amplitude change of the signal segment of each IMF component and the difference between the signal segment of each IMF component and the noise segment of each IMF component.
Specifically, for the firstThe (th) of the IMF components>The signal section is first obtained +.>The (th) of the IMF components>Variance and mean of all amplitudes in the signal section, combined with +.>All the amplitude averages in the noise section of the IMF component, obtain +.>Individual IMF componentsMiddle->The probability that each signal segment is a noise signal is specifically calculated by the following formula:
；
in the method, in the process of the invention,indicate->The (th) of the IMF components>The likelihood that the individual signal segments are noise signals; />Indicate->The (th) of the IMF components>Variance of all amplitudes in the individual signal segments; />Indicate->The (th) of the IMF components>The average of all amplitudes in the individual signal segments; />Indicate->All the amplitude averages in the noise section of the IMF components; />Representing a linear normalization function; />Representing an absolute value operation.
It should be further noted that,indicating->The (th) of the IMF components>Mean value of all amplitudes in the signal section and +.>Differences between all the amplitude averages in the noise segments of the IMF components; thus->The smaller the value of +.>The (th) of the IMF components>The more likely the individual signal segments are noise signals. Thus->The greater the value of +.>The (th) of the IMF components>The more likely the individual signal segments are noise signals.
So far, the likelihood that the signal segment in each IMF component is a noise signal is obtained.
Step S003: acquiring the isolation degree of the signal segment in each IMF component according to the possibility that the signal segment in each IMF component is a noise signal and the difference between the amplitudes; and acquiring the possibility that each sound signal segment is simultaneously talking by multiple people according to the isolation degree of the signal segment in each IMF component and the possibility that the signal segment in each IMF component is a noise signal.
It should be noted that, as a sound signal optimizing transmission method for a microphone, the present embodiment has a final purpose of improving the quality of the sound signal, and therefore, it is necessary to remove the noise signal from the sound signal, and since the conventional denoising method has a poor denoising effect when a plurality of persons speak simultaneously, it is necessary to obtain the possibility that a plurality of persons speak simultaneously for each time period. In order to obtain the possibility that multiple persons in each time period speak at the same time, it is first necessary to calculate the degree of isolation of the signal segments in each IMF component.
Specifically, for the calculation ofThe (th) of the IMF components>Isolation of individual signal segments is obtained by first obtaining the +.sup.th in each IMF component>The mean value of all amplitudes in each signal segment, the +.>The probability that the signal segment is a noise signal is based on the +.>The mean value of all amplitudes in each signal segment, the +.>The possibility that the signal segment is a noise signal, obtain +.>The (th) of the IMF components>The isolation degree of each signal segment is as follows:
；
in the method, in the process of the invention,indicate->The (th) of the IMF components>Isolation of individual signal segments; />Representing the number of IMF components;indicate->The (th) of the IMF components>The average of all amplitudes in the individual signal segments; />Indicate->The (th) of the IMF components>The average of all amplitudes in the individual signal segments; />Indicate->Of the IMF components/>The likelihood that the individual signal segments are noise signals; />Indicate->The (th) of the IMF components>The likelihood that the individual signal segments are noise signals; />Representing a linear normalization function; />Representing an absolute value operation.
It should be noted that the number of the substrates,indicating->The (th) of the IMF components>The average of all amplitudes in the signal segment, and +.>The (th) of the IMF components>The difference between the means of all amplitudes in the signal sections, therefore +.>The larger the value of (2), the description of +.>The (th) of the IMF components>Signal segments, and the first +.among all other IMF components>The greater the difference between the individual signal segments; />Indicating->The (th) of the IMF components>The probability that the signal segment is a noise signal, and +.>The (th) of the IMF components>The signal segments are the differences between the possibilities of noise signals, thus +.>The larger the value of (2), the description of +.>The (th) of the IMF components>Signal segment and->The (th) of the IMF components>The more unlikely the signal segments are together noise signals, or together not noise signals, i.e +.>The larger the value of (2), the description of +.>The (th) of the IMF components>Signal segments, and the first +.among all other IMF components>The greater the difference between the individual signal segments. Thus->The larger the value of (2) is, the descriptionThe (th) of the IMF components>Signal segments, and the first +.>The greater the difference in the individual signal segments.
It should be further noted that, when no person speaks, only a noise signal exists in the sound signal, so that the difference between all IMF signal segments is small in the period when no person speaks; when only one person speaks, noise signals and signals of one person speaking exist in the sound signals, so that in the period of time when only one person speaks, the difference among all IMF signal segments except the signal segment with the highest isolation degree is small; when a plurality of people speak simultaneously, noise signals and signals which are spoken by the plurality of people exist in the sound signals, so that in the period of time when the plurality of people speak simultaneously, the difference among all IMF signal segments is still large except for the signal segment with the highest isolation degree; on this basis, the possibility that each sound signal segment is in simultaneous talking by multiple persons can be obtained.
Specifically, for the calculation ofThe individual sound signal segments are at the possibility of multiple people speaking simultaneously; according to the +.>Isolation of individual signal segments, th +.>The possibility that the signal segment is a noise signal, obtain +.>The specific calculation formula of the sound signal segments is as follows:
；
in the method, in the process of the invention,indicate->The individual sound signal segments are at the possibility of multiple people speaking simultaneously; />Represents the +.o of all IMF components>The maximum value of the degree of isolation in the individual signal segments; />Represents the +.o of all IMF components>Minimum degree of isolation in individual signal segments; />Represents the +.o of all IMF components>The next highest value of isolation in the individual signal segments;represents the +.o of all IMF components>The maximum likelihood of a noise signal in the signal segments; />Representing a linear normalization function.
It should be further noted that, when the firstThe individual sound signal segments are in the absence of a person speaking,is small in value (I)>The value of (2) is large; when->The individual voice signal segments are in the sense that a person speaks, < >>Big value of->The value of (2) is small; when->The individual sound signal segments are in the case of simultaneous talking of multiple persons,/->Is small in value (I)>The value of (2) is small; thus->The greater the value of +.>Personal soundThe more likely the audio signal segment is in a period where multiple people are speaking simultaneously.
Thus, the possibility that each sound signal segment is simultaneously talking by multiple people is obtained.
Step S004: according to the possibility that each sound signal segment is simultaneously uttered by multiple people, adaptively generating the threshold value of wavelet threshold denoising of each sound signal segment; and denoising the sound signal according to the threshold value of wavelet threshold denoising of each sound signal segment.
It should be noted that, the present embodiment is used as a method for optimizing transmission of a sound signal for a microphone, and the final purpose is to improve the quality of the sound signal. After the possibility that each sound signal segment is simultaneously uttered by multiple persons is obtained in step S003, the threshold size of wavelet threshold denoising of each sound signal segment can be adaptively generated according to the possibility that each sound signal segment is simultaneously uttered by multiple persons.
Specifically, for adaptive generation of the firstThe threshold value of wavelet threshold denoising of each sound signal segment is first preset with an initial wavelet threshold +.>，/>The specific value of (1) can be set by combining with the actual situation, the hard requirement is not required in the embodiment, and +.>Description is made; based on +.>The degree of isolation of the individual signal segments, and the likelihood of each sound signal segment being in simultaneous speech by multiple persons, adaptively generates the +.>Wavelet of individual sound signal segmentsThe specific calculation formula of the threshold denoising threshold is as follows:
；
in the method, in the process of the invention,indicate->A wavelet threshold denoising threshold for each sound signal segment; />Representing the +.f among all IMF components>The maximum value of the degree of isolation in the individual signal segments; />Indicate->The individual sound signal segments are at the possibility of multiple people speaking simultaneously; />Representing a preset initial wavelet threshold; />Representing a linear normalization function.
It should be further noted that, in order to improve the quality of the sound signal, the present embodiment avoids removing the signal characteristics when the person speaks when using the wavelet threshold to perform denoising, so that the threshold of wavelet threshold denoising needs to be reduced when the person speaks, and the threshold of wavelet threshold denoising needs to be further reduced when the person speaks, so as to avoid the sound signal losing the signal characteristics when the person speaks.
Thus, the threshold size of wavelet threshold denoising of each sound signal segment is obtained. According to the threshold value of wavelet threshold denoising of each sound signal segment, performing wavelet threshold denoising processing on each sound signal segment, wherein the specific process of wavelet threshold denoising processing is a known prior art, so that redundant description is omitted in the embodiment, and denoising of sound signals is completed; the purpose of improving the quality of the sound signal is achieved.
This embodiment is completed.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the invention, but any modifications, equivalent substitutions, improvements, etc. within the principles of the present invention should be included in the scope of the present invention.
Claims (8)
1. A method for optimized transmission of sound signals for microphones, the method comprising the steps of:
collecting a sound signal through a microphone, and decomposing the sound signal to obtain a plurality of IMF components;
segmenting the sound signal and the IMF components to obtain a signal segment and a sound signal segment in each IMF component; obtaining a noise section of each IMF component according to the amplitude of each sound signal section and the time corresponding relation between each sound signal section and the signal section of each IMF component; acquiring the possibility that the signal segment in each IMF component is a noise signal according to the difference between each signal segment in each IMF component and the noise signal segment and the amplitude of each signal segment in each IMF component;
acquiring the isolation degree of the signal segment in each IMF component according to the possibility that the signal segment in each IMF component is a noise signal and the difference between the amplitudes; acquiring the possibility that each sound signal segment is simultaneously uttered by multiple people according to the isolation degree of the signal segment in each IMF component and the possibility that the signal segment in each IMF component is a noise signal;
according to the isolation degree of the signal segment in each IMF component and the possibility that the signal segment in each IMF component is a noise signal, the possibility that each sound signal segment is simultaneously uttered by multiple people is obtained, and the specific method comprises the following steps:
for calculation of the firstThe individual sound signal segments are at the possibility of multiple people speaking simultaneously; according to the +.>Isolation of individual signal segments, th +.>The possibility that the signal segment is a noise signal, obtain +.>The specific calculation formula of the sound signal segments is as follows:
；
in the method, in the process of the invention,indicate->The individual sound signal segments are at the possibility of multiple people speaking simultaneously; />Represents the +.o of all IMF components>The maximum value of the degree of isolation in the individual signal segments; />Represents the +.o of all IMF components>Minimum degree of isolation in individual signal segments; />Represents the +.o of all IMF components>The next highest value of isolation in the individual signal segments; />Represents the +.o of all IMF components>The maximum likelihood of a noise signal in the signal segments; />Representing a linear normalization function;
according to the possibility that each sound signal segment is simultaneously uttered by multiple people, adaptively generating the threshold value of wavelet threshold denoising of each sound signal segment;
the adaptive generation of the threshold size of wavelet threshold denoising of each sound signal segment according to the possibility that multiple people speak at the same time of each sound signal segment comprises the following specific methods:
for adaptive generation of the firstThe threshold value of wavelet threshold denoising of each sound signal segment is first preset with an initial wavelet threshold +.>The method comprises the steps of carrying out a first treatment on the surface of the Based on +.>The degree of isolation of the individual signal segments, and the likelihood of each sound signal segment being in simultaneous speech by multiple persons, adaptively generates the +.>The specific calculation formula of the wavelet threshold denoising threshold of each sound signal segment is as follows:
；
in the method, in the process of the invention,indicate->A wavelet threshold denoising threshold for each sound signal segment; />Representing the first of all IMF componentsThe maximum value of the degree of isolation in the individual signal segments; />Indicate->The individual sound signal segments are at the possibility of multiple people speaking simultaneously;representing a linear normalization function;
and denoising the sound signal according to the threshold value of wavelet threshold denoising of each sound signal segment.
2. The method for optimized transmission of sound signals of a microphone according to claim 1, wherein the method for acquiring sound signals by the microphone and decomposing the sound signals to obtain a plurality of IMF components comprises the following specific steps:
the sound signal is collected by a microphone and then decomposed by using an EMD algorithm to obtain a plurality of IMF components.
3. The method for optimized transmission of sound signals for microphones according to claim 1, wherein said segmenting sound signals and IMF components to obtain signal segments and sound signal segments in each IMF component comprises the following specific steps:
presetting a segment threshold valueDividing each IMF component signal and sound signal into sections of length +.>And obtaining the signal segment and the sound signal segment in each IMF component, and completing the segmentation of each IMF component and the segmentation of the sound signal.
4. The method for optimizing transmission of sound signals of a microphone according to claim 1, wherein the obtaining the noise segment of each IMF component according to the amplitude of each sound signal segment and the time correspondence between each sound signal segment and the signal segment of each IMF component comprises the following specific steps:
calculating all amplitude averages in each sound signal section, marking the sound signal section with the smallest amplitude average as a noise signal section, marking the time section corresponding to the noise signal section as a noise time section, and marking the IMF component signal section of each IMF component under the noise time section as the noise section of each IMF component.
5. The method for optimizing transmission of sound signals of a microphone according to claim 1, wherein the step of obtaining the probability that the signal segment in each IMF component is a noise signal according to the difference between each signal segment in each IMF component and the noise signal segment and the amplitude of each signal segment in each IMF component comprises the following specific steps:
for the followingFirst, theThe (th) of the IMF components>The signal section is first obtained +.>The (th) of the IMF components>Variance and mean of all amplitudes in the signal section, combined with +.>All the amplitude averages in the noise section of the IMF component, obtain +.>The (th) of the IMF components>The probability that each signal segment is a noise signal is specifically calculated by the following formula:
；
in the method, in the process of the invention,indicate->The (th) of the IMF components>The likelihood that the individual signal segments are noise signals; />Represent the first/>The (th) of the IMF components>Variance of all amplitudes in the individual signal segments; />Indicate->The (th) of the IMF components>The average of all amplitudes in the individual signal segments; />Indicate->All the amplitude averages in the noise section of the IMF components; />Representing a linear normalization function;representing an absolute value operation.
6. The method for optimizing transmission of sound signals of a microphone according to claim 1, wherein the obtaining the isolation degree of the signal segment in each IMF component according to the probability that the signal segment in each IMF component is a noise signal and the difference between the amplitudes comprises the following specific steps:
for calculation of the firstThe (th) of the IMF components>Isolation of individual signal segments is obtained by first obtaining the +.sup.th in each IMF component>The mean value of all amplitudes in each signal segment, the +.>The probability that the signal segment is a noise signal is based on the +.>The mean value of all amplitudes in each signal segment, the +.>The possibility that the signal segment is a noise signal, obtain +.>The (th) of the IMF components>Isolation of individual signal segments.
7. The method for optimized transmission of sound signals of microphone as claimed in claim 6, wherein said obtaining a firstThe (th) of the IMF components>The isolation degree of each signal segment comprises the following specific calculation formulas:
；
in the method, in the process of the invention,indicate->The (th) of the IMF components>Isolation of individual signal segments; />Representing the number of IMF components; />Indicate->The (th) of the IMF components>The average of all amplitudes in the individual signal segments; />Indicate->The (th) of the IMF components>The average of all amplitudes in the individual signal segments; />Indicate->The (th) of the IMF components>The likelihood that the individual signal segments are noise signals; />Indicate->The (th) of the IMF components>The likelihood that the individual signal segments are noise signals; />Representing a linear normalization function;representing an absolute value operation.
8. The method for optimized transmission of sound signals for microphones according to claim 1, wherein said denoising sound signals according to the threshold size of wavelet threshold denoising for each sound signal segment comprises the specific steps of:
and carrying out wavelet threshold denoising processing on each sound signal segment according to the threshold size of wavelet threshold denoising of each sound signal segment, and completing denoising of the sound signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311754560.1A CN117437931B (en) | 2023-12-20 | 2023-12-20 | Sound signal optimized transmission method for microphone |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311754560.1A CN117437931B (en) | 2023-12-20 | 2023-12-20 | Sound signal optimized transmission method for microphone |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117437931A CN117437931A (en) | 2024-01-23 |
| CN117437931B true CN117437931B (en) | 2024-04-02 |
Family
ID=89558590
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311754560.1A Active CN117437931B (en) | 2023-12-20 | 2023-12-20 | Sound signal optimized transmission method for microphone |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117437931B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105573104A (en) * | 2015-12-16 | 2016-05-11 | 上海大学 | Watch detection noise reduction method based on improved EMD |
| KR101842792B1 (en) * | 2016-10-05 | 2018-03-27 | 국방과학연구소 | An anti-jamming method and system based on Empirical Mode Decomposition and wavelet de-noising scheme |
| CN109344751A (en) * | 2018-09-20 | 2019-02-15 | 上海工程技术大学 | A kind of reconstructing method of internal car noise signal |
| CN113919389A (en) * | 2021-08-24 | 2022-01-11 | 国网浙江省电力有限公司电力科学研究院 | GIS fault diagnosis method and system based on voiceprint imaging |
-
2023
- 2023-12-20 CN CN202311754560.1A patent/CN117437931B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105573104A (en) * | 2015-12-16 | 2016-05-11 | 上海大学 | Watch detection noise reduction method based on improved EMD |
| KR101842792B1 (en) * | 2016-10-05 | 2018-03-27 | 국방과학연구소 | An anti-jamming method and system based on Empirical Mode Decomposition and wavelet de-noising scheme |
| CN109344751A (en) * | 2018-09-20 | 2019-02-15 | 上海工程技术大学 | A kind of reconstructing method of internal car noise signal |
| CN113919389A (en) * | 2021-08-24 | 2022-01-11 | 国网浙江省电力有限公司电力科学研究院 | GIS fault diagnosis method and system based on voiceprint imaging |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117437931A (en) | 2024-01-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9269368B2 (en) | Speaker-identification-assisted uplink speech processing systems and methods | |
| CN107393550B (en) | Voice processing method and device | |
| US8880396B1 (en) | Spectrum reconstruction for automatic speech recognition | |
| CN101930746B (en) | MP3 compressed domain audio self-adaptation noise reduction method | |
| Bharti et al. | A new spectral subtraction method for speech enhancement using adaptive noise estimation | |
| Ali et al. | Speech enhancement using dilated wave-u-net: an experimental analysis | |
| Alam et al. | Robust feature extraction for speech recognition by enhancing auditory spectrum | |
| JP4891805B2 (en) | Reverberation removal apparatus, dereverberation method, dereverberation program, recording medium | |
| CN117437931B (en) | Sound signal optimized transmission method for microphone | |
| Nie et al. | Deep Noise Tracking Network: A Hybrid Signal Processing/Deep Learning Approach to Speech Enhancement. | |
| Hassani et al. | Speech enhancement based on spectral subtraction in wavelet domain | |
| Li et al. | Joint sparse representation based cepstral-domain dereverberation for distant-talking speech recognition | |
| Shao et al. | A versatile speech enhancement system based on perceptual wavelet denoising | |
| Ghanbari et al. | Spectral subtraction in the wavelet domain for speech enhancement | |
| Meutzner et al. | Binaural signal processing for enhanced speech recognition robustness in complex listening environments | |
| CN118016079B (en) | Intelligent voice transcription method and system | |
| Vanajakshi et al. | Modified Wiener filter based speech restoration with autocorrelation of isolated Kannada word speech recognition | |
| CN117711419B (en) | Intelligent data cleaning method for data center | |
| CN110070874A (en) | A kind of voice de-noising method and device for Application on Voiceprint Recognition | |
| Gao et al. | DNN Speech Separation Algorithm Based on Improved Segmented Masking Target | |
| Wang et al. | Speech recognition using blind source separation and dereverberation method for mixed sound of speech and music | |
| Aravinda et al. | Digital Preservation and Noise Reduction using Machine Learning | |
| CN116913308A (en) | Single-channel voice enhancement method for balancing noise reduction amount and voice quality | |
| Ghasemi | A new approach based on SVD for speech enhancement | |
| CN108281154B (en) | Noise reduction method for voice signal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |