CN115798502A - Audio denoising method for Bluetooth headset - Google Patents
Audio denoising method for Bluetooth headset Download PDFInfo
- Publication number
- CN115798502A CN115798502A CN202310043255.5A CN202310043255A CN115798502A CN 115798502 A CN115798502 A CN 115798502A CN 202310043255 A CN202310043255 A CN 202310043255A CN 115798502 A CN115798502 A CN 115798502A
- Authority
- CN
- China
- Prior art keywords
- audio
- value
- deviation value
- fitting
- obtaining
- 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.)
- Granted
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The invention relates to the technical field of voice processing, in particular to an audio denoising method for a Bluetooth headset. The method comprises the steps of obtaining a trend term curve by carrying out amplitude fitting on an obtained audio signal, dividing audio paragraphs according to the trend term curve, obtaining a fitting deviation value according to amplitude difference in the audio paragraphs, analyzing difference of the fitting deviation value of the audio paragraphs to obtain a first sampling density, analyzing probability density distribution conditions of different types of fitting deviation values of the audio paragraphs in the whole audio signal to obtain a second sampling density, obtaining the number of sampling points of the audio paragraphs according to sampling density coefficients obtained by the first sampling density and the second sampling density, and sampling and denoising the audio according to the number of the sampling points. According to the invention, through analyzing the audio data, the self-adaption increase based on the minimum number of sampling points is realized when each audio paragraph is put in the sampling points, the number of the sampling points is reduced to the greatest extent, and the quality and the efficiency of audio denoising are improved.
Description
Technical Field
The invention relates to the technical field of voice processing, in particular to an audio denoising method for a Bluetooth headset.
Background
Although the bluetooth headset is convenient to carry and use, the bluetooth headset also has the disadvantages that the transmission performance of the bluetooth headset is always weaker than that of wired transmission no matter how advanced the bluetooth technology is, the problems of audio data loss, damage, noise and the like often occur in the transmission process, audio data distortion is caused, and a decoding chip of a mobile phone also generates slight noise in the process of analyzing the audio. And if the environment of the user is complex, for example, strong electromagnetic interference exists outside, even the loud noise of the bluetooth headset occurs. The influence on the external environment cannot be completely avoided, but the problem of distortion of audio data can be reduced by optimizing the audio transmission mode, and because the signal value of an audio signal can be accurately determined at any time in the time sequence, the existing wired and wireless transmission of the audio is realized by using a sampling coding mode, so that the method for improving the audio quality mostly considers how to scientifically and reasonably adjust the distribution density of sampling points during coding transmission, not only ensures the signal restoration degree, but also can control the transmission time to reduce noise interference.
The quality of the audio signal depends to some extent on the magnitude of the sampling frequency, and the higher the sampling frequency, the higher the signal restorability, but the transmission time of the audio data is correspondingly increased, and the longer the transmission time, the more intensive the noise interference. In the existing method for denoising audio at sampling points, the sampling points are mostly analyzed to reduce the number of the sampling points, so as to remove redundant noise points, and the audio processing method does not reduce the number of the sampling points to the maximum extent, so that the simplified coding efficiency is reduced, the transmission time is increased, and the noise attachment rate is higher during transmission.
Disclosure of Invention
In order to solve the technical problems that the number of sampling points is not reduced to the maximum extent during audio processing, so that the reduction of coding efficiency is reduced, the transmission time is prolonged, and the noise attachment rate is high during transmission in the prior art, the invention aims to provide an audio denoising method for a Bluetooth headset, and the adopted technical scheme is as follows:
the invention provides an audio denoising method for a Bluetooth headset, which comprises the following steps:
acquiring an audio signal, and dividing audio paragraphs according to a trend term curve obtained by audio signal amplitude fitting; obtaining a fitting deviation value of each audio paragraph according to the audio signal and the trend term curve;
obtaining a deviation value set consisting of the fitting deviation values in each audio paragraph according to the distribution of the fitting deviation values, and obtaining a first sampling density of each audio paragraph according to the difference of the fitting deviation values in the deviation value set;
obtaining probability density distribution of each type of fitting deviation value according to the type density distribution condition of all fitting deviation values in the audio signal; taking any one type of fitting deviation value in the deviation value set of any one audio paragraph as a reference fitting deviation value, obtaining a type influence value of the reference fitting deviation value through the proportion of the reference fitting deviation value and the corresponding probability density distribution, and obtaining a second sampling density of the audio paragraph according to the type influence values of the fitting deviation values of all the types;
obtaining a sampling density coefficient according to the first sampling density and the second sampling density, and obtaining the number of sampling points of each audio paragraph according to the ratio of each audio paragraph in the audio signal and the sampling density coefficient; and sampling and denoising the audio according to the number of the sampling points.
Further, the obtaining of the set of bias values includes:
and arranging the fitting deviation values in the audio paragraph according to a normal distribution mode to obtain a deviation value set and a serial number of each fitting deviation value in the deviation value set.
Further, the obtaining of the first sampling density comprises:
obtaining a maximum fitting deviation value in the deviation value set and a corresponding maximum serial number; in the deviation value set, obtaining a minimum fitting deviation value on the left side of the maximum fitting deviation value, recording the minimum fitting deviation value as a left minimum fitting deviation value, recording a sequence number corresponding to the left minimum fitting deviation value as a left minimum sequence number, obtaining a minimum fitting deviation value on the right side of the maximum fitting deviation value, recording a sequence number corresponding to the right minimum fitting deviation value as a right minimum sequence number;
taking the average value of the left minimum fitting deviation value and the right minimum fitting deviation value as an average minimum deviation value, and taking the difference value of the maximum fitting deviation value and the average minimum deviation value as a deviation range value; taking the absolute value of the difference between the maximum sequence number and the left minimum sequence number as a left sequence difference, taking the absolute value of the difference between the maximum sequence number and the right minimum sequence number as a right sequence difference, and taking the average value of the left sequence difference and the right sequence difference as a sequence difference value;
and taking the ratio of the deviation value to the sequence difference value as a first sampling density.
Further, the obtaining of the probability density distribution includes:
counting the types and the number of all fitting deviation values in the audio signal to obtain a distribution probability histogram of the fitting deviation values, and obtaining the type distribution probability of each type of fitting deviation value according to the distribution probability histogram; and taking the result obtained by taking the probability of each type distribution as the independent variable of the Gaussian probability density function as the probability density distribution of the corresponding class fitting deviation value.
Further, the obtaining of the type influence value includes:
taking the result of the fixed integral of the probability density distribution corresponding to the reference fitting deviation value as a first reference value;
and taking the ratio of the number of the reference fitting deviation values in the audio paragraphs to the total number of the reference fitting deviation values in the audio signals as a number ratio, and multiplying the number ratio by the first reference value to obtain the type influence value of the reference fitting deviation values.
Further, the obtaining of the second sampling density comprises:
normalizing the type influence value of each type of the fit deviation value in the audio paragraph to obtain a corresponding normalized type influence value, and taking the addition result of the normalized type influence values as the second sampling density of the audio paragraph.
Further, the obtaining of the sampling density coefficient includes:
and taking the average value of the first sampling density and the second sampling density as a sampling density coefficient.
Further, the obtaining of the number of sample points of each audio paragraph includes:
obtaining the lowest sampling point number of the audio signal, and multiplying the lowest sampling point number by the length ratio of the audio paragraph in the audio signal for any audio paragraph to obtain the paragraph lowest sampling point number of the audio paragraph; and obtaining sampling distribution weight according to the sampling density coefficient, and multiplying the lowest sampling point number of the paragraphs with the sampling distribution weight to obtain the sampling point number of the audio paragraphs.
Further, the obtaining of the audio passage comprises:
taking the amplitude of the audio signal as a data point, and performing curve fitting by adopting a least square method to obtain a trend term curve; and obtaining extreme values in the trend term curve, dividing a time sequence section between every two adjacent extreme values into an audio section, and obtaining all audio sections in the audio signal.
Further, the obtaining of the fitting deviation value comprises:
and obtaining a signal amplitude value of each moment in the audio signal, obtaining a fitting numerical value of each moment on the trend term curve, and taking a difference value between the signal amplitude value corresponding to each moment and the fitting numerical value as a fitting deviation value.
The invention has the following beneficial effects:
the method mainly divides the obtained audio paragraphs through the trend term curve of the amplitude fitting of the audio signal, wherein the fitting deviation value in each audio paragraph can embody the discrete characteristics of the audio signal on the corresponding time sequence of the paragraph. The discrete features reflect the distribution of the sound features in the audio paragraph, and when the discreteness is larger, the sound features of the audio paragraph are more dense, and the number of samples required is larger. Therefore, the difference of the fitting deviation values in each audio paragraph and the probability density distribution of different types of fitting deviation values in each audio paragraph in the whole audio signal are analyzed, and the sampling density coefficient of the audio paragraph, namely the retention weight of the sampling point, is obtained through comprehensive calculation, so that the transmission of the retained audio signal is completed. The number of the sampling points is increased in a self-adaptive mode according to the density coefficient of the sampling points of each audio paragraph based on the number of the sampling points, the number of the sampling points is reduced to the maximum extent, the integrity of the audio is guaranteed, the sampling efficiency is greatly improved, the transmission time is shortened, and the noise attachment probability in transmission is reduced.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of an audio denoising method for a bluetooth headset according to an embodiment of the present invention.
Detailed Description
To further illustrate the technical means and effects of the present invention for achieving the predetermined objects, the following detailed description of an audio denoising method for a bluetooth headset according to the present invention with reference to the accompanying drawings and preferred embodiments shows the following detailed descriptions of the embodiments, structures, features and effects thereof. In the following description, the different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
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 describes a specific scheme of an audio denoising method for a bluetooth headset according to the present invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of an audio denoising method for a bluetooth headset according to an embodiment of the present invention is shown, where the method includes:
step S1: acquiring an audio signal, and dividing audio paragraphs according to a trend term curve obtained by audio signal amplitude fitting; and obtaining the fitting deviation value of each audio paragraph according to the audio signal and the trend term curve.
In the embodiment of the invention, the audio denoising method of the Bluetooth headset mainly aims to improve the quality of audio data in wireless transmission, the quality of an audio signal depends on the size of sampling frequency to a certain extent, the higher the sampling frequency is, the higher the signal restoring degree is, but the transmission time of the audio data can be correspondingly increased, and the longer the transmission time is, the higher the attachment rate of noise is, and the denser the noise interference is. Therefore, the embodiment of the invention focuses on optimizing the number of coded samples of audio data during wireless transmission, and the wireless transmission has the disadvantages of poor anti-interference capability and increased noise attachment probability due to redundant sampling points of an audio file, so that the sampling points are selected from the lowest sampling point number according to the characteristic requirement of each audio paragraph, the audio integrity is ensured, and meanwhile, the audio transmission time and the noise generated by electromagnetic interference in the transmission process are reduced to the greatest extent. Therefore, when an audio signal is analyzed, firstly, an audio paragraph needs to be obtained, then, the requirement characteristics of sampling points are increased according to the fitting deviation value analysis of each audio paragraph, so that the audio signal is obtained, the audio paragraphs are divided according to the trend term curve obtained by the audio signal amplitude fitting, and the fitting deviation value of each audio paragraph is obtained according to the audio signal and the trend term curve, which specifically comprises the following steps:
in the embodiment of the present invention, a piece of audio data is intercepted from a multimedia device, and since the nature of sound is a wave transmitted by vibration, and sound can be regarded as a series of analog signals when being transmitted through an audio line or a microphone, the recording form of the audio signal is a waveform diagram continuous in time sequence, wherein the vertical axis of the waveform diagram is the amplitude of sound wave energy, the horizontal axis is time, and the number of sound waves in unit time is frequency. It should be noted that the obtaining of the audio signal is well known to those skilled in the art, and is not described herein.
The amplitude of the audio signal is used as a data point, a least square method is adopted to perform curve fitting to obtain a trend term curve, and the obtained trend term curve can represent the characteristics of sound tones of sound in a section of audio, such as the high, low, slow and fast time sequence, and the like, so that the smoothness degree of the trend term curve needs to be considered when a sampling point is selected. It should be noted that the least squares fitting is a well-known technique in the art and will not be described herein.
Because the trend term curve represents the sound characteristics, the audio paragraphs can be divided according to the trend term curve, so that each audio paragraph can be analyzed subsequently, and the number of sampling points of each audio paragraph can be obtained. The method comprises the steps of firstly obtaining extrema in a trend term curve, dividing a time sequence section between every two adjacent extrema into an audio frequency section, and obtaining all audio frequency sections in an audio frequency signal, wherein each audio frequency section has monotonicity. In the embodiment of the present invention, derivation is performed on the trend term curve to obtain an extreme value, and a point where the derivative is zero is used as an extreme value point.
More or less sound characteristics may exist in each divided audio paragraph, and the fitting deviation value at the moment is obtained through the amplitude of the original audio signal and the corresponding numerical value on the trend term curve, which specifically includes:
obtaining the signal amplitude of each moment in the audio signal, obtaining the fitting numerical value of each moment on the trend term curve, taking the difference value between the signal amplitude corresponding to each moment and the fitting numerical value as a fitting deviation value, wherein the expression of the fitting deviation value is as follows:
in the formula (I), the compound is shown in the specification,
expressed as a deviation value of the fit,
is shown as being at
The amplitude of the audio signal at the time of day,
expressed as the first on the trend term curve
The fitted value of the moment.
The difference between the trend term curve and the amplitude of the audio signal is obtained by using a difference value method, when the difference is larger, the fitting deviation value is larger, the sound characteristic which is possessed at the corresponding moment is more obvious, meanwhile, for the whole audio paragraph, the fitting deviation value can reflect the dispersion of the audio data on the time sequence of the paragraph, the dispersion reflects the distribution condition of the sound characteristic, when the dispersion of the data is larger, the sound characteristic in the audio paragraph is more, more sampling points need to be set at the moment, the dispersion is smaller, the distribution of the sound characteristic is less, and the number of the sampling points is also less at the moment.
Step S2: and obtaining a deviation value set consisting of the fitting deviation values in each audio paragraph according to the distribution of the fitting deviation values, and obtaining the first sampling density of each audio paragraph according to the difference of the fitting deviation values in the deviation value set.
According to the step S1, through the analysis of the fitting deviation value, the sound feature distribution condition in the audio paragraphs can be obtained, more sampling points are added to the audio paragraphs with more sound features to ensure the integrity of the audio, and the audio paragraphs with less sound feature distribution do not need redundant sampling points to retain the signal features, so when analyzing the fitting deviation value, the difference of the fitting deviation value in each audio paragraph is analyzed first, a deviation value set composed of the fitting deviation value in each audio paragraph is obtained according to the distribution of the fitting deviation value, and the first sampling density of each audio paragraph is obtained according to the difference of the fitting deviation value in the deviation value set, which specifically includes:
according to the fluctuation characteristics of the audio signal, for the audio paragraph with the fitting deviation value, the audio paragraph has the position with the larger fitting deviation value, and the distribution of the fitting deviation value has the characteristic of Gaussian normal, namely, the amplitude position corresponding to the audio signal and the adjacent amplitude thereof are in a normal distribution shape with a high middle and low two sides. Therefore, in order to better analyze the whole audio paragraph, the fitting deviation values in the audio paragraph are all arranged according to a normal distribution mode, that is, the fitting deviation values sequentially decrease from the middle maximum value to the two ends, so as to obtain a deviation value set and a sequence number corresponding to each fitting deviation value in the deviation value set.
The whole analysis can be performed on each audio paragraph according to the deviation value set, and when the difference of the fitting deviation value is larger, that is, the variation trend is larger, it indicates that the audio paragraph needs more sampling points, so the first sampling density of each audio paragraph is obtained according to the deviation value set, and the obtaining of the specific first sampling density includes:
and acquiring the maximum fitting deviation value and the corresponding maximum serial number in the deviation value set, comprehensively analyzing the two sides of the maximum fitting deviation value in the deviation value set according to the characteristic of normal distribution of the fitting deviation value, and comprehensively judging the trend changes of the two sides. And obtaining the minimum fitting deviation value on the left side of the maximum fitting deviation value, recording the minimum fitting deviation value as a left minimum fitting deviation value, recording the sequence number corresponding to the left minimum fitting deviation value as a left minimum sequence number, obtaining the minimum fitting deviation value on the right side of the maximum fitting deviation value, recording the sequence number corresponding to the right minimum fitting deviation value as a right minimum sequence number.
And respectively obtaining a deviation range value and a sequence difference value of the fitting deviation value, obtaining the overall change trend condition of the fitting deviation value through comprehensive analysis of two sides, taking the average value of the left minimum fitting deviation value and the right minimum fitting deviation value as an average minimum deviation value, and taking the difference value of the maximum fitting deviation value and the average minimum deviation value as the deviation range value. And taking the absolute value of the difference value between the maximum sequence number and the left minimum sequence number as a left sequence difference, taking the absolute value of the difference value between the maximum sequence number and the right minimum sequence number as a right sequence difference, and taking the average value of the left sequence difference and the right sequence difference as a sequence difference value.
Taking the ratio of the deviation range value to the sequence difference value as a first sampling density, in the embodiment of the present invention, for the accuracy of the subsequent calculation, the expression of the first sampling density is:
in the formula (I), the compound is shown in the specification,
a first sampling density expressed as an audio paragraph;
expressed as the maximum fitted deviation value in the set of deviation values;
the maximum serial number corresponding to the maximum fitting deviation value is represented;
the minimum fitting deviation value represented as the left side of the maximum fitting deviation value, namely the left minimum fitting deviation value;
the left minimum serial number corresponding to the left minimum fitting deviation value is represented;
represented as the left side of the maximum fit deviation value in the set of deviation values;
the minimum fitting deviation value on the right side of the maximum fitting deviation value is expressed, namely the right minimum fitting deviation value;
representing the right minimum serial number corresponding to the right minimum fitting deviation value;
shown as the right side of the maximum fit deviation value in the set of deviation values.
Comprehensively analyzing the variation trends on both sides of the maximum fitting deviation value in normal distribution by using a ratio form,
expressed as the average obtained by averaging the left and right minimum fitted deviation valuesThe minimum deviation value of the deviation value is set,
expressed as deviation range value, the range of the deviation value set is obtained by integrating the two sides of the maximum fitting deviation value,
indicated as the difference in the left sequence,
indicated as a difference in the right sequence,
and expressing the sequence difference values obtained by averaging the left sequence difference and the right sequence difference, and synthesizing the maximum fitting deviation value to obtain the sequence difference corresponding to the deviation value set and reflecting the change trend. The ratio of the deviation threshold value to the sequence difference value represents the variation trend of the fitting deviation value of each audio paragraph, namely the first sampling density, when the variation trend of the fitting deviation value is larger, and then the sound feature distribution of the corresponding audio paragraph is more, at this moment, the sampling density needs to be increased, namely the number of sampling points is increased.
And step S3: obtaining the probability density distribution of each type of fitting deviation value according to the type density distribution condition of all fitting deviation values in the audio signal; and taking any one type of fitting deviation value in the deviation value set of any one audio paragraph as a reference fitting deviation value, obtaining the type influence value of the reference fitting deviation value through the proportion of the reference fitting deviation value and the corresponding probability density distribution, and obtaining the second sampling density of the audio paragraph according to the type influence values of the fitting deviation values of all the types.
In step S2, the difference analysis of the fitting deviation values of each audio segment is completed, and further, the probability density distribution of the different types of fitting deviation values in each audio segment in the whole audio signal is analyzed. In the audio signal, because the fitting deviation values of different types have different occurrence probabilities and the fitting deviation values of different types in each audio paragraph have different influences on the distribution probability of the fitting deviation values of the whole audio signal, the probability densities of the fitting deviation values of different types are analyzed, and the second sampling density is obtained according to the influence of the fitting deviation values of different types in each audio paragraph on the whole audio signal. Therefore, first, obtaining probability density distribution conditions corresponding to the fitting deviation values of each type in the audio signal, and obtaining probability density distribution of each type of fitting deviation value according to the type density distribution conditions of all the fitting deviation values in the audio signal specifically includes:
preferably, the types and the number of all fitting deviation values in the audio signal are counted to obtain a distribution probability histogram of the fitting deviation values, and the type distribution probability of each type of fitting deviation value is obtained according to the distribution probability histogram, wherein each type distribution probability represents the distribution probability condition of the fitting deviation value of the corresponding type. In order to further more accurately represent the influence of each type of fitting deviation value in each audio paragraph, a result obtained by taking each type distribution probability as an independent variable of a gaussian probability density function is taken as the probability density distribution of the corresponding type of fitting deviation value, and a probability density distribution value reflects the probability density of the type distribution probability of the corresponding type of fitting deviation value in the gaussian probability density function.
In the formula (I), the compound is shown in the specification,
is shown as the first
The probability density distribution of the type distribution probability of the class-fit deviation value in the Gaussian probability density function is
，
Is shown as the first
The class fits the type distribution probability of the bias value,
expressed as the standard deviation of the type distribution probability of the fitted deviation value,
expressed as the mean of the type distribution probabilities of the fitted deviation values,
expressed as a circumferential ratio of the number of revolutions,
expressed as an exponential function with a natural constant as the base. It should be noted that the application of the gaussian probability density function formula is a technical means well known to those skilled in the art, and therefore, the meaning of the specific formula is not described in detail.
After obtaining the probability density distribution of all types of fitting deviation values, analyzing each audio paragraph, selecting different numbers of sampling points for the audio paragraphs according to different influences of the fitting deviation value of each type on the distribution probability in the whole audio signal, so that any type of fitting deviation value in the deviation value set of any audio paragraph is taken as a reference fitting deviation value, obtaining the type influence value of the reference fitting deviation value through the proportion of the reference fitting deviation value and the corresponding probability density distribution, and obtaining the second sampling density of the audio paragraph according to the type influence value of the fitting deviation value of all types, which specifically comprises:
because each class of fitting deviation value in each audio paragraph needs to be analyzed, any class of fitting deviation value in the deviation value set of any audio paragraph is taken as a reference fitting deviation value, and the probability density distribution corresponding to the reference fitting deviation value is taken as a result of fixed integration as a first reference value, where the first reference value reflects the possibility that the distribution probability of the reference fitting deviation value in the audio paragraph is the type distribution probability, that is, the expected value of the type distribution probability of the reference fitting deviation value in the gaussian probability density function.
Taking the ratio of the number of the reference fitting deviation value in the audio paragraph to the total number of the reference fitting deviation value in the audio signal as a number ratio, wherein the number ratio reflects the ratio of the reference fitting deviation value in the audio paragraph to the reference deviation value in the entire audio signal, and when the ratio of the reference fitting deviation value in the audio paragraph is larger, the influence degree of the reference fitting deviation value of the audio paragraph is larger, and the number ratio is multiplied by the first reference value to obtain a type influence value of the reference fitting deviation value.
In the formula (I), the compound is shown in the specification,
is shown as
Second of an audio paragraph
The class fits the type impact value of the bias value,
is shown as
A segment of the audio frequency is displayed,
is shown as
The class fitting deviation value is also referred to as a reference fitting deviation value,
is shown as the first
The first in the audio passage
The number of class-fit deviation values,
is shown as
The number of class fit deviation values in the overall audio signal,
is shown as
The probability density distribution of the type distribution probability of the class-fit deviation value in the Gaussian probability density function is
，
Is shown as
The class fits the type distribution probability of the bias value,
is shown as to
The type distribution probability of the class-fit deviation value takes the first reference value of the integral result, i.e. the first reference value
The class fits the expected value of the type distribution probability of the deviation value in the gaussian probability density function.
Comprehensively analyzing the quantity ratio and the first reference value in the form of product,
is shown as the first
The first in the audio passage
Class fitting deviation value is in the whole audio signal
The larger the number ratio is, the larger the influence of the reference fitting deviation value in the audio paragraph on the whole audio signal is, and therefore, the larger the type influence value is; when the first reference value is larger, the description is given in the second
The more likely the distribution probability of the reference fitting deviation value in an audio segment is the type distribution probability, and the distribution of the reference fitting deviation value of the audio segment is close to the distribution of the reference fitting deviation value in the whole audio signal, which shows that the second time
The influence of the reference fit deviation value in an audio segment is large, so the type influence value will be larger.
The type influence value represents the influence of the reference fitting deviation value in the audio paragraph on the distribution probability of the reference fitting deviation value in the whole audio signal, when the type influence value is larger, the reference fitting deviation value in the audio paragraph has a large influence on the reference fitting deviation value in the whole audio signal, and a sampling point needs to be added to the audio paragraph, so that the whole audio signal is more complete, and the information retention is more accurate.
Obtaining type influence values of fitting deviation values of all classes in an audio paragraph, normalizing the type influence values of the fitting deviation values of all classes in the audio paragraph to obtain corresponding normalized type influence values, taking an addition result of the normalized type influence values as a second sampling density of the audio paragraph, wherein the second sampling density integrates the influence of the fitting deviation values of all the classes in each audio paragraph, and a sampling point needs to be added to the audio paragraph with large influence to ensure the integrity of the whole audio signal.
In the formula (I), the compound is shown in the specification,
is shown as
The second sampling density of the individual audio segments,
is shown as
The values of the deviation are fitted to the classes,
expressed as the total number of types of deviation values fitted,
is shown as
Second of an audio paragraph
The class fits the type impact value of the bias value,
the function is expressed as a hyperbolic tangent function, and it should be noted that the hyperbolic tangent function is a technical means well known to those skilled in the art, and is not described herein in detail.
Comprehensively analyzing all class fitting deviation values of the audio paragraph in an accumulation mode, and utilizing
Hyperbolic tangent function will be
Normalizing the type influence value of the class-fit deviation value, wherein the normalization processing is used for limiting the value range of the numerical value and normalizingAdding the type influence values of all the normalized class fit deviation values, taking the obtained accumulated value as a second sampling density of the audio paragraph, and when the accumulated value is larger, the larger the distribution influence of the fit deviation values of all the types in the whole audio signal is, the larger the distribution influence of the audio paragraph on the whole audio signal is, the larger the influence of the audio paragraph on the whole audio signal is, and the more the audio paragraph needs to be paid attention to ensure the integrity of the whole audio signal.
And step S4: obtaining a sampling density coefficient according to the first sampling density and the second sampling density, and obtaining the number of sampling points of each audio paragraph according to the ratio of each audio paragraph in the audio signal and the sampling density coefficient; and sampling and denoising the audio according to the number of the sampling points.
According to the step S2 and the step S3, the difference of the fitting deviation value in each audio paragraph and the influence of different types of fitting deviation values in each audio paragraph on the whole audio signal are respectively analyzed, the first sampling density and the second sampling density of each audio paragraph are respectively obtained, and finally, the final sampling density coefficient is obtained through integration. At this time, the final sampling density coefficient can reflect the degree of the number of sampling points required to be added to each audio paragraph, the theoretical minimum number of sampling points of each audio paragraph is obtained through calculation, and the number of the sampling points is adaptively increased according to the sampling density coefficient based on the minimum number of the sampling points, so that the integrity of the audio signal is ensured. Therefore, obtaining a sampling density coefficient according to the first sampling density and the second sampling density, and obtaining the number of sampling points of each audio paragraph according to the ratio of each audio paragraph in the audio signal and the sampling density coefficient specifically includes:
taking the average value of the obtained first sampling density and the second sampling density of each audio section as a final sampling density coefficient, wherein the sampling density coefficient expression is as follows:
in the formula (I), the compound is shown in the specification,
expressed as tonesThe sampling density coefficient of the frequency segment is,
expressed as a first sampling density of the audio passage,
represented as a second sampling density of the audio passage.
The first sampling density and the second sampling density are integrated in a mean value mode, and when the first sampling density and the second sampling density are larger, the larger the first sampling density and the second sampling density are, the larger the sampling density coefficient is, which indicates that the audio paragraph needs to add more sampling points to ensure the integrity of the audio signal.
When the audio signal is stored and transmitted, the encoding process needs to be carried out firstly, the common encoding sampling rate cannot be lower than half of the audio signal, and the higher the sampling rate is, the more the acquisition quantity per second is, the closer the sampling points are, the smoother the obtained audio curve is, and the more natural the transition is. However, a higher sampling rate also means that the audio file storage capacity is larger, the transmission time is longer, and the possibility of noise interference is higher. Therefore, when the sampling points are selected, the lowest sampling point number of the whole audio signal, namely half of the number of waves in the whole original audio signal, is obtained first.
Analyzing any one audio paragraph, multiplying the lowest sampling point number by the length ratio of the audio paragraph in the audio signal to obtain the paragraph lowest sampling point number of the audio paragraph, wherein the paragraph lowest sampling point number of each audio paragraph represents the release number of the theoretical sampling point in each audio paragraph, and increasing the number of the sampling points according to the corresponding sampling density coefficient on the basis of the paragraph lowest sampling point number. According to the embodiment of the invention, in consideration of the accuracy of calculation, the acquisition expression of the number of the audio paragraph sampling points is as follows:
in the formula (I), the compound is shown in the specification,
is shown as
The number of sample points of a segment of audio,
expressed as the length of the entire audio signal,
expressed as the number of original signal waves of the entire audio signal,
is shown as
The length of a segment of audio is,
is shown as
The sample density coefficient of an individual audio segment,
expressed as the lowest number of sample points,
is shown as
The length of an audio segment is a proportion of the length of the entire audio signal,
denoted as sample distribution weights.
Increasing sampling points in the lowest sampling point number of each audio paragraph according to different sampling distribution weights in a self-adaptive manner, finally obtaining the number of the sampling points after the sampling points are increased in a self-adaptive manner on the release number of the theoretical sampling points,
expressed as the lowest sampling point of the paragraph, namely the release number of the theoretical sampling point of the audio paragraph,
expressed as sample distribution weights, in which
The weighting is expressed as the number of sampling points to be increased according to the lowest number of sampling points of the paragraphs, the number is increased according to different influences of each audio paragraph, and it is to be noted that the data types corresponding to the number of sampling points are integers, so that the number of sampling points obtained by final calculation is rounded down to obtain the number of real sampling points of each audio paragraph.
And sampling the audio signal according to the number of the sampling points and carrying out denoising treatment to finish denoising. In the embodiment of the invention, the audio data obtained by sampling is quantized, coded and transmitted by adopting a PCM coding method. It should be noted that PCM coding is a technical means well known to those skilled in the art, and is not described herein.
So far, the sampling step of the whole audio signal is completed, the self-adaptive sampling point method considers the characteristics of the signal and the integrity of different audio paragraph signals, based on the number of minimum standard sampling points, the self-adaptive increase is carried out according to the requirements of different audio paragraph signals, the integrity of the signal is ensured, the coding efficiency of the audio is higher, unnecessary sampling points are reduced, the number of the sampling points is reduced to the maximum extent, the transmission interaction time is reduced, the noise attachment rate during transmission is reduced, and the quality and the efficiency of audio denoising are improved.
In summary, the present invention obtains an audio signal, obtains a trend term curve according to the fitting of the audio signal amplitude to divide audio paragraphs, obtains a fitting deviation value according to the audio signal and the trend term curve, obtains a first sampling density of each audio paragraph by analyzing the difference of the fitting deviation value for each audio paragraph, obtains a second sampling density of each audio paragraph by analyzing the probability density distribution of different types of fitting deviation values in the audio paragraph in the whole audio signal for each audio paragraph, obtains a sampling density coefficient according to the first sampling density and the second sampling density, obtains the number of sampling points of each audio paragraph according to the ratio and the sampling density coefficient of each audio paragraph in the audio signal, and samples and denoises the audio according to the number of the sampling points. According to the invention, through analyzing the audio data, the characteristics of the signal and the integrity of different audio paragraph signals are considered, and the self-adaptive increase is carried out according to the requirements of different audio paragraph signals based on the minimum standard sampling point number.
It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. The processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.
Claims (10)
1. An audio denoising method for a Bluetooth headset, the method comprising:
acquiring an audio signal, and dividing audio paragraphs according to a trend term curve obtained by audio signal amplitude fitting; obtaining a fitting deviation value of each audio paragraph according to the audio signal and the trend term curve;
obtaining a deviation value set consisting of the fitting deviation values in each audio paragraph according to the distribution of the fitting deviation values, and obtaining a first sampling density of each audio paragraph according to the difference of the fitting deviation values in the deviation value set;
obtaining the probability density distribution of each type of fitting deviation value according to the type density distribution condition of all fitting deviation values in the audio signal; taking any one type of fitting deviation value in the deviation value set of any one audio paragraph as a reference fitting deviation value, obtaining a type influence value of the reference fitting deviation value through the proportion of the reference fitting deviation value and the corresponding probability density distribution, and obtaining a second sampling density of the audio paragraph according to the type influence values of the fitting deviation values of all the types;
obtaining a sampling density coefficient according to the first sampling density and the second sampling density, and obtaining the number of sampling points of each audio paragraph according to the ratio of each audio paragraph in the audio signal and the sampling density coefficient; and sampling and denoising the audio according to the number of the sampling points.
2. The method of claim 1, wherein the obtaining of the set of bias values comprises:
and arranging the fitting deviation values in the audio paragraph according to a normal distribution mode to obtain a deviation value set and a serial number of each fitting deviation value in the deviation value set.
3. The method of claim 2, wherein the obtaining of the first sampling density comprises:
obtaining a maximum fitting deviation value in the deviation value set and a corresponding maximum serial number; in the deviation value set, obtaining a minimum fitting deviation value on the left side of the maximum fitting deviation value, recording the minimum fitting deviation value as a left minimum fitting deviation value, recording a sequence number corresponding to the left minimum fitting deviation value as a left minimum sequence number, obtaining a minimum fitting deviation value on the right side of the maximum fitting deviation value, recording a sequence number corresponding to the right minimum fitting deviation value as a right minimum sequence number;
taking the average value of the left minimum fitting deviation value and the right minimum fitting deviation value as an average minimum deviation value, and taking the difference value of the maximum fitting deviation value and the average minimum deviation value as a deviation range value; taking the absolute value of the difference value between the maximum sequence number and the left minimum sequence number as a left sequence difference, taking the absolute value of the difference value between the maximum sequence number and the right minimum sequence number as a right sequence difference, and taking the average value of the left sequence difference and the right sequence difference as a sequence difference value;
and taking the ratio of the deviation value to the sequence difference value as a first sampling density.
4. The method of claim 1, wherein the obtaining of the probability density distribution comprises:
counting the types and the number of all fitting deviation values in the audio signal to obtain a distribution probability histogram of the fitting deviation values, and obtaining the type distribution probability of each type of fitting deviation value according to the distribution probability histogram; and taking the result obtained by taking the probability of each type distribution as the independent variable of the Gaussian probability density function as the probability density distribution of the corresponding class fitting deviation value.
5. The audio denoising method for a bluetooth headset according to claim 1, wherein the obtaining of the type impact value comprises:
taking the result of the fixed integral of the probability density distribution corresponding to the reference fitting deviation value as a first reference value;
and taking the ratio of the number of the reference fitting deviation values in the audio paragraphs to the total number of the reference fitting deviation values in the audio signals as a number ratio, and multiplying the number ratio by the first reference value to obtain the type influence value of the reference fitting deviation values.
6. The method of claim 1, wherein the obtaining of the second sampling density comprises:
normalizing the type influence value of each type of the simulated deviation value in the audio paragraph to obtain a corresponding normalized type influence value, and taking the addition result of the normalized type influence values as the second sampling density of the audio paragraph.
7. The method of claim 1, wherein the obtaining of the sampling density coefficient comprises:
and taking the average value of the first sampling density and the second sampling density as a sampling density coefficient.
8. The method of claim 1, wherein the obtaining of the number of sample points per audio segment comprises:
obtaining the lowest sampling point number of the audio signal, and multiplying the lowest sampling point number by the length ratio of the audio paragraph in the audio signal for any audio paragraph to obtain the paragraph lowest sampling point number of the audio paragraph; and obtaining sampling distribution weight according to the sampling density coefficient, and multiplying the lowest sampling point number of the paragraphs with the sampling distribution weight to obtain the sampling point number of the audio paragraphs.
9. The method of claim 1, wherein the obtaining of the audio passage comprises:
taking the amplitude of the audio signal as a data point, and performing curve fitting by adopting a least square method to obtain a trend term curve; and obtaining extreme values in the trend term curve, dividing a time sequence section between every two adjacent extreme values into an audio section, and obtaining all audio sections in the audio signal.
10. The method of claim 1, wherein the obtaining of the fitting deviation value comprises:
and obtaining a signal amplitude of each moment in the audio signal, obtaining a fitting numerical value of each moment on the trend term curve, and taking a difference value between the signal amplitude corresponding to each moment and the fitting numerical value as a fitting deviation value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310043255.5A CN115798502B (en) | 2023-01-29 | 2023-01-29 | Audio denoising method for Bluetooth headset |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310043255.5A CN115798502B (en) | 2023-01-29 | 2023-01-29 | Audio denoising method for Bluetooth headset |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115798502A true CN115798502A (en) | 2023-03-14 |
| CN115798502B CN115798502B (en) | 2023-04-25 |
Family
ID=85429057
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310043255.5A Active CN115798502B (en) | 2023-01-29 | 2023-01-29 | Audio denoising method for Bluetooth headset |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115798502B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116405823A (en) * | 2023-06-01 | 2023-07-07 | 深圳市匠心原创科技有限公司 | Intelligent audio denoising enhancement method for bone conduction earphone |
| CN116991199A (en) * | 2023-09-25 | 2023-11-03 | 深圳扬兴科技有限公司 | Clock signal intelligent noise reduction method for clock chip |
| CN117357094A (en) * | 2023-11-15 | 2024-01-09 | 读取(北京)科技有限公司 | Respiratory state verification system using sound intensity and carbon dioxide detection |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0386820A1 (en) * | 1989-03-03 | 1990-09-12 | Koninklijke Philips Electronics N.V. | Method and arrangement for probabilistic pitch measuring |
| US20080312913A1 (en) * | 2005-04-01 | 2008-12-18 | National Institute of Advanced Industrial Sceince And Technology | Pitch-Estimation Method and System, and Pitch-Estimation Program |
| US20140142878A1 (en) * | 2012-11-21 | 2014-05-22 | Horizon Analog, Inc. | Density function centric signal processing |
| CN105117537A (en) * | 2015-08-13 | 2015-12-02 | 电子科技大学 | Weight comparison based resampling method for particle filter system |
| CN109102818A (en) * | 2018-08-29 | 2018-12-28 | 昆明理工大学 | A kind of denoising audio sample algorithm based on signal frequency probability density function profiles |
| US20190222188A1 (en) * | 2016-10-07 | 2019-07-18 | Sony Corporation | Information processing device, information processing method, and program |
-
2023
- 2023-01-29 CN CN202310043255.5A patent/CN115798502B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0386820A1 (en) * | 1989-03-03 | 1990-09-12 | Koninklijke Philips Electronics N.V. | Method and arrangement for probabilistic pitch measuring |
| US20080312913A1 (en) * | 2005-04-01 | 2008-12-18 | National Institute of Advanced Industrial Sceince And Technology | Pitch-Estimation Method and System, and Pitch-Estimation Program |
| US20140142878A1 (en) * | 2012-11-21 | 2014-05-22 | Horizon Analog, Inc. | Density function centric signal processing |
| CN105117537A (en) * | 2015-08-13 | 2015-12-02 | 电子科技大学 | Weight comparison based resampling method for particle filter system |
| US20190222188A1 (en) * | 2016-10-07 | 2019-07-18 | Sony Corporation | Information processing device, information processing method, and program |
| CN109102818A (en) * | 2018-08-29 | 2018-12-28 | 昆明理工大学 | A kind of denoising audio sample algorithm based on signal frequency probability density function profiles |
Non-Patent Citations (2)
| Title |
|---|
| 汤显峰;葛泉波;: "基于子带无迹粒子滤波的语音增强算法" * |
| 郭小华;陈文岳;: "基于压缩感知的资源节约型音频信号采集方法" * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116405823A (en) * | 2023-06-01 | 2023-07-07 | 深圳市匠心原创科技有限公司 | Intelligent audio denoising enhancement method for bone conduction earphone |
| CN116405823B (en) * | 2023-06-01 | 2023-08-29 | 深圳市匠心原创科技有限公司 | Intelligent audio denoising enhancement method for bone conduction earphone |
| CN116991199A (en) * | 2023-09-25 | 2023-11-03 | 深圳扬兴科技有限公司 | Clock signal intelligent noise reduction method for clock chip |
| CN116991199B (en) * | 2023-09-25 | 2024-01-02 | 深圳扬兴科技有限公司 | Clock signal intelligent noise reduction method for clock chip |
| CN117357094A (en) * | 2023-11-15 | 2024-01-09 | 读取(北京)科技有限公司 | Respiratory state verification system using sound intensity and carbon dioxide detection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115798502B (en) | 2023-04-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115798502A (en) | Audio denoising method for Bluetooth headset | |
| JP5089772B2 (en) | Apparatus and method for detecting voice activity | |
| KR100896737B1 (en) | Device and method for robustry classifying audio signals, method for establishing and operating audio signal database and a computer program | |
| US20210074312A1 (en) | Method and Apparatus for Detecting a Voice Activity in an Input Audio Signal | |
| CN102918588B (en) | A spatial audio processor and a method for providing spatial parameters based on an acoustic input signal | |
| CN1727860B (en) | Noise suppression method and apparatus | |
| KR102296680B1 (en) | Audio signal classification method and device | |
| US20110081026A1 (en) | Suppressing noise in an audio signal | |
| JP5127754B2 (en) | Signal processing device | |
| CN110265065B (en) | Method for constructing voice endpoint detection model and voice endpoint detection system | |
| CN103247298B (en) | A kind of sensitivity correction method and audio frequency apparatus | |
| CN111312290B (en) | Audio data tone quality detection method and device | |
| CN112004177B (en) | Howling detection method, microphone volume adjustment method and storage medium | |
| CN103325384A (en) | Harmonicity estimation, audio classification, pitch definition and noise estimation | |
| JP2012522255A (en) | Audio signal classification method and apparatus | |
| CN103544957B (en) | Method and device for bit distribution of sound signal | |
| CN104981870B (en) | Sound enhancing devices | |
| US20090037172A1 (en) | Method for generating a vector codebook, method and device for compressing data, and distributed speech recognition system | |
| CN113113049A (en) | Voice activity detection method combined with voice enhancement | |
| CN113327589A (en) | Voice activity detection method based on attitude sensor | |
| CN115472153A (en) | Voice enhancement system, method, device and equipment | |
| CN117542370B (en) | Intelligent voice signal processing method and system for intelligent customer service robot | |
| EP4193357A1 (en) | Maintaining invariance of sensory dissonance and sound localization cues in audio codecs | |
| CN117198312B (en) | Voice interaction processing method for intelligent glasses | |
| Srinivasarao | An efficient recurrent Rats function network (Rrfn) based speech enhancement through noise reduction |
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 |