CN115575896B - Feature enhancement method for non-point sound source image - Google Patents

Feature enhancement method for non-point sound source image Download PDF

Info

Publication number
CN115575896B
CN115575896B CN202211524057.2A CN202211524057A CN115575896B CN 115575896 B CN115575896 B CN 115575896B CN 202211524057 A CN202211524057 A CN 202211524057A CN 115575896 B CN115575896 B CN 115575896B
Authority
CN
China
Prior art keywords
sound source
sound
array
point
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
Application number
CN202211524057.2A
Other languages
Chinese (zh)
Other versions
CN115575896A (en
Inventor
曹祖杨
杜子哲
陶慧芳
侯佩佩
张凯强
洪全付
张永全
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Crysound Electronics Co Ltd
Original Assignee
Hangzhou Crysound Electronics Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hangzhou Crysound Electronics Co Ltd filed Critical Hangzhou Crysound Electronics Co Ltd
Priority to CN202211524057.2A priority Critical patent/CN115575896B/en
Publication of CN115575896A publication Critical patent/CN115575896A/en
Application granted granted Critical
Publication of CN115575896B publication Critical patent/CN115575896B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
    • G01S5/18—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using ultrasonic, sonic, or infrasonic waves
    • G01S5/20—Position of source determined by a plurality of spaced direction-finders

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Circuit For Audible Band Transducer (AREA)

Abstract

The invention relates to a characteristic enhancement method for a non-point sound source image, which comprises the following steps: s1, determining the sound source position of a non-point sound source; s2, obtaining an array focusing sound source signal; s3, obtaining a plurality of sound source frequency spectrum characteristics; s4, obtaining each frequency domain signal received by each array element; s5 is in
Figure 100004_DEST_PATH_IMAGE001
Forming time signal points corresponding to all times in the dimensional space; s6, clustering the signal points at each moment to form
Figure DEST_PATH_IMAGE002
Of individual, i.e. non-point, sound sources
Figure 557803DEST_PATH_IMAGE002
A sound emitting area; s7, obtaining each sounding region from
Figure 185093DEST_PATH_IMAGE001
A fitted straight line formed in a dimensional space; s8, calculating to obtain the regional spectrum characteristics of each sounding area based on the filtering signals received by each array element at the final moment and the weight coefficients of the signals received by each array element in each sounding area; and S9, carrying out secondary imaging on the basis of the frequency spectrum characteristics of each region, the positions of the array elements and the sound source position of the non-point sound source to obtain a sound source image with enhanced characteristics. The invention can enhance the characteristics of the sound source image of the non-point sound source, thereby improving the imaging effect of the sound imaging instrument on the weak sound source.

Description

Feature enhancement method for non-point sound source image
Technical Field
The invention belongs to the technical field of sound source positioning, and particularly relates to a feature enhancement method for a non-point sound source image.
Background
Acoustic imaging (acoustic imaging) is based on a microphone array measurement technology, and is characterized in that the position of a sound source is determined according to a phased array principle by measuring the phase difference of signals of sound waves in a certain space reaching each microphone, the amplitude of the sound source is measured, and the distribution of the sound source in the space is displayed in an image mode, namely a cloud image-sound image of the spatial sound field distribution is obtained, wherein the intensity is represented by the color and the brightness of the image.
For example, chinese patent publication No. CN110082725a discloses a sound source localization delay estimation method and a sound source localization system based on a microphone array, which integrate two improved frequency domain weighting functions, i.e., PATH and ML, by using a newly proposed frequency domain weighting function, and make up for the disadvantage that the original algorithm cannot resist noise and reverberation at the same time. Firstly, a microphone array receives two paths of signals, the two paths of signals are converted into digital signals through ADC sampling, windowing and framing are carried out on the two paths of signals, then frequency domain signals are obtained through Fourier transformation, cross power spectrums and weighting functions of the two frames of signals are calculated, weighting is carried out on the cross power spectrums, then cross correlation functions of the two paths of signals are obtained through inverse Fourier transformation on the weighted cross power spectrums, and finally peak detection is carried out on the cross correlation functions to obtain the relative time delay of the two paths of signals. The method reduces the influence of the environmental noise and reverberation on the time delay estimation, improves the accuracy of the time delay estimation and improves the sound source positioning precision.
For another example, chinese patent publication No. CN113126028a discloses a noise source positioning method based on multiple microphone arrays. M microphone sensors are selected to construct an annular microphone array, one microphone sensor is arranged to serve as a reference microphone sensor, an array coordinate system is established by the reference microphone sensor, the other M-1 microphone sensors are arranged around the reference microphone sensor, and D sound sources are arranged in a cabin; obtaining relative transfer functions from D sound sources to each microphone sensor, and constructing an array flow pattern matrix of the annular microphone array; further introducing the linear distance between the sound source and the reference microphone sensor, the azimuth angle of the sound source relative to the reference microphone sensor and the sound source frequency to construct an array flow pattern near-field model; estimating the azimuth angle of each sound source relative to the reference microphone sensor by adopting a MUSIC algorithm; more than two identical annular microphone arrays are preset in the cabin, the azimuth angle of the sound source relative to each annular microphone array relative to the reference microphone sensor is estimated, the distance from the sound source to each annular microphone array is solved by using a least square method overall, and then the sound source position information is obtained.
Therefore, at present, the research on sound source positioning is mature, but the imaging research on the sound image is less, and when the acoustic imaging instrument images a weak sound source, the final imaging effect is poor due to weak sound source signals, so that the effect of finally displaying the image in front of a user is poor. Therefore, a method for enhancing the features of the sound source image is needed.
Disclosure of Invention
In view of the above problems in the prior art, the present invention provides a method for enhancing characteristics of a sound source image of a non-point sound source, which can enhance characteristics of the sound source image of the non-point sound source, thereby improving an imaging effect of a sound imaging apparatus on a weak sound source. The invention adopts the following technical scheme:
a feature enhancement method for a non-point sound source image comprises the following steps:
s1, determining the sound source position of a non-point sound source;
s2, pointing the microphone array of the acoustic imaging instrument to the sound source position to obtain an array focusing sound source signal;
s3, carrying out frequency spectrum search on the array focusing sound source signals to obtain a plurality of sound source frequency spectrum characteristics;
s4, performing band-pass filtering processing on the sound source signals received by the array elements according to a plurality of sound source frequency spectrum characteristics to obtain filtering signals received by the array elements, and performing frequency domain processing on the filtering signals to obtain frequency domain signals;
s5, according to the frequency domain signals received by all array elements at each moment
Figure DEST_PATH_IMAGE001
Forming a time signal point corresponding to each time in the dimensional space,
Figure 186412DEST_PATH_IMAGE002
time signal point corresponding to time
Figure DEST_PATH_IMAGE003
Has the coordinates of
Figure 259410DEST_PATH_IMAGE004
，
Figure DEST_PATH_IMAGE005
To represent
Figure 837153DEST_PATH_IMAGE002
At time first
Figure 464444DEST_PATH_IMAGE006
The frequency domain signal received by an array element, i.e. at
Figure 707206DEST_PATH_IMAGE001
In a dimensional space of
Figure DEST_PATH_IMAGE007
The coordinates of the individual dimensions are such that,
Figure 385443DEST_PATH_IMAGE001
representing the total number of microphone array elements,
Figure 55459DEST_PATH_IMAGE008
，
Figure DEST_PATH_IMAGE009
represents the final time;
s6, clustering the signal points at each moment based on a clustering algorithm to form
Figure 983095DEST_PATH_IMAGE010
Of individual, i.e. non-point, sound sources
Figure 763969DEST_PATH_IMAGE010
A sound emitting area;
s7, fitting the time signal points in each sound-emitting area to obtain each sound-emitting area
Figure 280401DEST_PATH_IMAGE001
The slope of the fitting straight line in different dimensions represents the weight coefficient of each array element in the sound production area corresponding to the fitting straight line;
s8, calculating to obtain sound source signals of all sounding areas at the final moment based on the filtering signals received by all array elements at the final moment and the weight coefficients of the signals received by all array elements at all sounding areas, and obtaining the area spectrum characteristics of all sounding areas based on the sound source signals of all sounding areas;
and S9, carrying out secondary imaging based on the frequency spectrum characteristics of each region, the positions of each array element and the sound source position of the non-point sound source to obtain a sound source image with enhanced characteristics, and carrying out sound source positioning.
Preferably, step S3 includes the following steps:
s3.1, carrying out differential operation on the array focusing sound source signals:
Figure DEST_PATH_IMAGE011
，
wherein,
Figure 423717DEST_PATH_IMAGE012
representing the differentiated array focused sound source signal,
Figure DEST_PATH_IMAGE013
representing the array focused sound source signal,
Figure 291179DEST_PATH_IMAGE014
means differentiation;
s3.2, carrying out frequency spectrum search on the differentiated array focused sound source signals to obtain a plurality of sound source frequency spectrum characteristics
Figure DEST_PATH_IMAGE015
And each sound source frequency spectrum characteristic meets the following conditions:
Figure 157634DEST_PATH_IMAGE016
，
Figure DEST_PATH_IMAGE017
，
wherein,
Figure 59731DEST_PATH_IMAGE018
denotes the first
Figure DEST_PATH_IMAGE019
The spectral characteristics of the individual sound sources are,
Figure 150178DEST_PATH_IMAGE020
representing an imaging threshold.
Preferably, the method further comprises the following steps between step S5 and step S6:
normalizing the signal points at each moment to obtain normalized signal points corresponding to each moment, wherein the normalized calculation formula is as follows:
Figure DEST_PATH_IMAGE021
，
Figure 255669DEST_PATH_IMAGE022
，
wherein,
Figure DEST_PATH_IMAGE023
to represent
Figure 175083DEST_PATH_IMAGE024
The frequency domain signal received by the 1 st array element at the moment,
Figure DEST_PATH_IMAGE025
to represent
Figure 210648DEST_PATH_IMAGE024
A standardized signal point corresponding to the moment;
in step S6, the standardized signal points corresponding to each time are clustered based on a clustering algorithm to form
Figure 393367DEST_PATH_IMAGE010
Of individual, i.e. non-point, sound sources
Figure 969842DEST_PATH_IMAGE010
And a sound emitting area.
Preferably, step S6 includes the following steps:
s6.1, initialization
Figure 646942DEST_PATH_IMAGE010
A cluster center;
s6.2, calculating the distance between each standardized signal point and each clustering center, and dividing each standardized signal point into the class to which the clustering center closest to the standardized signal point belongs;
s6.3, based on all the standardized signal points in each class, recalculating to obtain the iterated signal
Figure 726894DEST_PATH_IMAGE010
A cluster center;
s6.4, repeating the step S6.2 to the step S6.3 until the preset iteration times are reached to obtain the final product
Figure 80515DEST_PATH_IMAGE010
Of individual, i.e. non-point, sound sources
Figure 762862DEST_PATH_IMAGE010
And a sound emitting area.
Preferably, in step S7, the slopes of the different dimensions of each of the fitting straight lines formed in the dimensional space in each of the voicing regions can be represented as follows:
Figure 243653DEST_PATH_IMAGE026
in which
Figure DEST_PATH_IMAGE027
Is shown as
Figure 709269DEST_PATH_IMAGE028
The fitting straight line corresponding to each sound production area is
Figure 499371DEST_PATH_IMAGE007
Slope of one dimension, i.e. representing the second
Figure 532661DEST_PATH_IMAGE007
Array element is in
Figure 597569DEST_PATH_IMAGE028
The individual voicing regions receive the weighting coefficients of the signal.
Preferably, in step S7, the time signal points in each sound emission region are fitted by using the least square method to obtain each sound emission region
Figure 652113DEST_PATH_IMAGE001
A fitted straight line formed in a dimensional space.
Preferably, in step S8, the sound source signal of each sound emission area at the final time is calculated based on the following equation:
Figure DEST_PATH_IMAGE029
，
wherein,
Figure 895007DEST_PATH_IMAGE030
indicates the last moment
Figure 402211DEST_PATH_IMAGE001
The filtered signals received by the individual array elements,
Figure DEST_PATH_IMAGE031
is shown as
Figure 490384DEST_PATH_IMAGE001
The array element is in
Figure 665013DEST_PATH_IMAGE028
The individual voicing regions receive the weighting coefficients of the signal,
Figure 328076DEST_PATH_IMAGE032
indicates the first time at the final moment
Figure 338888DEST_PATH_IMAGE028
Sound source signals of the individual sound emitting areas.
Preferably, step S9 includes the following steps:
s9.1, calculating an array flow pattern of each array element at each area spectrum characteristic position based on the area spectrum characteristics of each sound production area of the non-point sound source, the position of each array element and the sound source position;
and S9.2, performing secondary imaging based on the array flow pattern of each array element at each region spectrum feature to obtain a feature-enhanced sound source image.
Preferably, in step S9.1, the calculation formula of the array flow pattern of each array element at each region spectral feature is as follows:
Figure DEST_PATH_IMAGE033
，
wherein,
Figure 480020DEST_PATH_IMAGE034
is shown as
Figure DEST_PATH_IMAGE035
The regional spectral characteristics of the individual sound-emanating regions,
Figure 522538DEST_PATH_IMAGE036
is shown as
Figure 559764DEST_PATH_IMAGE007
Array element is in
Figure 307140DEST_PATH_IMAGE035
The array flow pattern at the spectral features of the individual sound emanating areas,
Figure DEST_PATH_IMAGE037
a position of a sound source is indicated,
Figure 799432DEST_PATH_IMAGE038
the number of the imaginary numbers is represented by,
Figure DEST_PATH_IMAGE039
which is indicative of the speed of sound,
Figure 214233DEST_PATH_IMAGE040
is shown as
Figure 438672DEST_PATH_IMAGE007
Coordinates of individual array elements.
Preferably, in step S9.2, the calculation formula of the feature-enhanced sound source image is:
Figure DEST_PATH_IMAGE041
，
wherein,
Figure 938924DEST_PATH_IMAGE042
representing a feature-enhanced sound source image,
Figure DEST_PATH_IMAGE043
is shown as
Figure 907011DEST_PATH_IMAGE044
The regional spectral characteristics of the individual sound-emanating regions,
Figure DEST_PATH_IMAGE045
is shown as
Figure 441897DEST_PATH_IMAGE007
The area spectrum characteristics in the sound source signal received by each array element are
Figure 820926DEST_PATH_IMAGE046
The composition of (1).
The invention has the beneficial effects that:
the characteristic enhancement can be carried out on the sound source image of the non-point sound source, and the imaging effect of the sound imaging instrument on the sound source is further improved.
Because the mixing frequencies of the sounding components at different positions of the non-point sound source can be different, and a dominant sound source exists at different moments, the signal correlation of the sound source frequency needs to be searched, and the imaging enhancement is respectively performed on different positions of the non-point sound source to achieve the image enhancement effect.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in 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 for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a flowchart of a feature enhancement method for a non-point sound source image according to the present invention.
Detailed Description
The following description of the embodiments of the present invention is provided by way of specific examples, and other advantages and effects of the present invention will be readily apparent to those skilled in the art from the disclosure herein. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.
Referring to fig. 1, the present embodiment provides a feature enhancement method for a non-point sound source image, including the steps of:
s1, determining the sound source position of a non-point sound source;
s2, pointing the microphone array of the acoustic imaging instrument to the sound source position to obtain an array focusing sound source signal;
s3, carrying out frequency spectrum search on the array focusing sound source signals to obtain a plurality of sound source frequency spectrum characteristics;
s4, performing band-pass filtering processing on the sound source signals received by the array elements according to a plurality of sound source frequency spectrum characteristics to obtain filtering signals received by the array elements, and performing frequency domain processing on the filtering signals to obtain frequency domain signals;
s5, according to the frequency domain signals received by all array elements at each moment
Figure 556276DEST_PATH_IMAGE001
Forming a time signal point corresponding to each time in the dimensional space,
Figure 311743DEST_PATH_IMAGE002
time signal point corresponding to time
Figure 169977DEST_PATH_IMAGE003
Has the coordinates of
Figure 736219DEST_PATH_IMAGE004
，
Figure 679904DEST_PATH_IMAGE005
To represent
Figure 770220DEST_PATH_IMAGE002
At time first
Figure 233694DEST_PATH_IMAGE006
The frequency domain signal received by an array element, i.e. at
Figure 220104DEST_PATH_IMAGE001
In a dimensional space
Figure 651085DEST_PATH_IMAGE007
The coordinates of the individual dimensions are such that,
Figure 748354DEST_PATH_IMAGE001
representing the total number of elements of the microphone array,
Figure 66334DEST_PATH_IMAGE008
，
Figure 958067DEST_PATH_IMAGE009
represents the final time;
it should be noted that, the above-mentioned dimensions may refer to a two-dimensional coordinate system and a three-dimensional coordinate system, where the coordinates of the two-dimensional coordinate system are (x, y), and the coordinates of the three-dimensional coordinate system are (x, y, z), where x is the coordinates representing the first dimension in the two-dimensional coordinate system and the three-dimensional coordinate system, y is the coordinates representing the second dimension in the two-dimensional coordinate system and the three-dimensional coordinate system, and z is the coordinates representing the third dimension in the three-dimensional coordinate system.
S6, clustering the signal points at all the moments based on a clustering algorithm to form
Figure 876345DEST_PATH_IMAGE010
Of individual, i.e. non-point, sound sources
Figure 511725DEST_PATH_IMAGE010
A sound emitting area;
s7, fitting the time signal points in each sound-emitting area to obtain each sound-emitting area
Figure 946861DEST_PATH_IMAGE001
A fitted straight line formed in a dimensional space, the fitted straight line being differentThe slope of the dimension represents the weight coefficient of each array element for receiving signals in the sounding area corresponding to the fitting straight line, and the least square method is adopted for fitting in the embodiment;
s8, calculating to obtain sound source signals of all sounding areas at the final moment based on the filtering signals received by all array elements at the final moment and the weight coefficients of the signals received by all array elements at all sounding areas, and obtaining the area spectrum characteristics of all sounding areas based on the sound source signals of all sounding areas;
and S9, carrying out secondary imaging based on the frequency spectrum characteristics of each region, the positions of each array element and the sound source position of the non-point sound source to obtain a sound source image with enhanced characteristics.
It should be noted that: although a point sound source exists in an ideal state and is relatively few in real life, a sound source with a small sound emitting area can be approximated to a point sound source, and the non-point sound source in this embodiment represents a sound source with a large sound emitting area (the sound emitting area can also be regarded as a diaphragm).
Because the mixing frequencies of the sounding components at different positions of the non-point sound source can be different, and a dominant sound source exists at different moments, the signal correlation of the sound source frequency needs to be searched, and the imaging enhancement is respectively performed on different positions of the non-point sound source to achieve the effect of image enhancement.
Therefore, the invention can enhance the characteristics of the sound source image of the non-point sound source, thereby improving the imaging effect of the acoustic imaging instrument on the weak sound source.
Specifically, the method comprises the following steps:
in step S1, the output result of the acoustic imager is a two-dimensional image, and the physical meaning is that the stronger the energy of the sound source is, the brighter the sound source position in the image is, the horizontal scanning angle is the abscissa and the vertical scanning angle is the ordinate in the image, so that the position of the sound source can be determined by energy peak search, and the sound source position is recorded as
Figure 275074DEST_PATH_IMAGE037
。
In step S2, the acoustic imaging instrument microphone array has
Figure DEST_PATH_IMAGE047
An array element, the first
Figure 493697DEST_PATH_IMAGE006
The signals received by an array element are recorded as
Figure 198348DEST_PATH_IMAGE048
The frequency domain signal of each array element can be obtained using fast fourier transform:
Figure DEST_PATH_IMAGE049
，
wherein,
Figure 5767DEST_PATH_IMAGE050
the frequency is represented by a frequency-dependent signal,
Figure DEST_PATH_IMAGE051
representing a fast fourier transform operation.
It is known that the sound source position is calculated by step S1
Figure 583510DEST_PATH_IMAGE037
So the pan microphone array points to the sound source location to get an array focused sound source signal:
Figure 210800DEST_PATH_IMAGE052
，
Figure DEST_PATH_IMAGE053
，
wherein,
Figure 631328DEST_PATH_IMAGE038
the number of the imaginary numbers is represented by,
Figure 840723DEST_PATH_IMAGE040
denotes the first
Figure 510739DEST_PATH_IMAGE007
The coordinates of the individual array elements are,
Figure 625326DEST_PATH_IMAGE039
representing the speed of sound.
In step S3, the method includes the following steps:
s3.1, carrying out differential operation on the array focusing sound source signals:
Figure 406200DEST_PATH_IMAGE011
，
wherein,
Figure 673364DEST_PATH_IMAGE012
representing the differentiated array focused sound source signal,
Figure 514281DEST_PATH_IMAGE013
representing the array focused sound source signal,
Figure 647322DEST_PATH_IMAGE014
means differentiation;
s3.2, carrying out frequency spectrum search on the differentiated array focusing sound source signals to obtain a plurality of sound source frequency spectrum characteristics
Figure 982620DEST_PATH_IMAGE015
And each sound source frequency spectrum characteristic meets the following conditions:
Figure 353558DEST_PATH_IMAGE016
，
Figure 365377DEST_PATH_IMAGE054
，
wherein,
Figure 188976DEST_PATH_IMAGE018
denotes the first
Figure 325035DEST_PATH_IMAGE019
The spectral characteristics of the individual sound sources are,
Figure 284901DEST_PATH_IMAGE020
which represents the imaging threshold, which in this embodiment is the average of the spectral characteristics of a plurality of sound sources.
The method also comprises the following steps between the step S5 and the step S6:
normalizing the signal points at each moment to obtain normalized signal points corresponding to each moment, wherein the normalized calculation formula is as follows:
Figure DEST_PATH_IMAGE055
，
Figure 998779DEST_PATH_IMAGE056
，
wherein,
Figure 325986DEST_PATH_IMAGE023
to represent
Figure 252354DEST_PATH_IMAGE024
The frequency domain signal received by the 1 st array element at the moment,
Figure DEST_PATH_IMAGE057
to represent
Figure 827911DEST_PATH_IMAGE024
A standardized signal point corresponding to the moment;
in step S6, the standardized signal points corresponding to each time are clustered based on a clustering algorithm to form
Figure 994581DEST_PATH_IMAGE010
Of individual, i.e. non-point, sound sources
Figure 792773DEST_PATH_IMAGE010
And a sound emitting area.
Step S6 includes the following steps:
s6.1, initialization
Figure 257252DEST_PATH_IMAGE010
A cluster center;
s6.2, calculating the distance between each standardized signal point and each clustering center, and dividing each standardized signal point into the class to which the clustering center closest to the standardized signal point belongs;
s6.3, based on all the standardized signal points in each class, recalculating to obtain the iterated signal
Figure 457289DEST_PATH_IMAGE010
A cluster center;
s6.4, repeating the step S6.2 to the step S6.3 until the preset iteration times are reached to obtain the final product
Figure 998123DEST_PATH_IMAGE010
Of individual, i.e. non-point, sound sources
Figure 752453DEST_PATH_IMAGE010
And a sound emitting area. The preset iteration number can be set according to the actual situation, and the iteration is carried out until
Figure 286202DEST_PATH_IMAGE010
The individual cluster centers are not changed.
In step S7, the slopes of different dimensions of each of the sound-emitting areas, which are the fitted straight lines formed in the dimensional space, can be represented as:
Figure 340746DEST_PATH_IMAGE058
wherein
Figure 580710DEST_PATH_IMAGE027
Is shown as
Figure 353494DEST_PATH_IMAGE028
The fitting straight line corresponding to each sound production area is
Figure 425355DEST_PATH_IMAGE007
Slope of one dimension, i.e. representing the second
Figure 819558DEST_PATH_IMAGE007
The array element is in
Figure 951462DEST_PATH_IMAGE028
The individual voicing regions receive the weighting coefficients of the signal.
Why the following fitting straight line formed in the pair-dimensional space exists
Figure 945963DEST_PATH_IMAGE001
The individual slopes are interpreted:
taking a two-dimensional plane as an example, if there is a straight line in the two-dimensional plane, in the front view and the top view of the two-dimensional plane, there are two slopes for the straight line, that is, there is a slope for each of the x-dimension and the y-dimension.
Taking a three-dimensional space as an example, if a straight line exists in the three-dimensional space, in the front view, the top view and the side view of the three-dimensional space, the straight line has three slopes, that is, an x-dimension, a y-dimension and a z-dimension each have one slope.
This fitted straight line thus formed in the dimensional space, from the different dimensions of which there should be a line which is present
Figure 821515DEST_PATH_IMAGE001
A slope.
In step S8, the sound source signal of each sound emission area at the final time is calculated based on the following equation:
Figure DEST_PATH_IMAGE059
，
wherein,
Figure 132542DEST_PATH_IMAGE030
indicates the first time at the final moment
Figure 966506DEST_PATH_IMAGE001
Array element connectorThe received filtered signal is then transmitted to the receiver,
Figure 464614DEST_PATH_IMAGE031
denotes the first
Figure 878278DEST_PATH_IMAGE001
Array element is in
Figure 761921DEST_PATH_IMAGE028
The individual voicing regions receive the weighting coefficients of the signal,
Figure 970048DEST_PATH_IMAGE032
indicates the last moment
Figure 952523DEST_PATH_IMAGE028
Sound source signals of the individual sound emitting areas.
The above formula is written as a set of equations that can be expressed as:
Figure 169878DEST_PATH_IMAGE060
，
can be written into simultaneously
Figure 908027DEST_PATH_IMAGE001
Linear form under dimensional space:
Figure DEST_PATH_IMAGE061
，
therefore it is first
Figure 631263DEST_PATH_IMAGE007
Array element is in
Figure 87652DEST_PATH_IMAGE028
The weight coefficient and the second of the received signal of each sounding region
Figure 328272DEST_PATH_IMAGE028
The fitting straight line corresponding to each sound production area is
Figure 186506DEST_PATH_IMAGE007
The slopes of the dimensions are equal.
In step S9, the method includes the following steps:
s9.1, calculating an array flow pattern of each array element at each area spectrum characteristic position based on the area spectrum characteristics of each sounding area of the non-point sound source, the position of each array element and the position of the sound source;
and S9.2, performing secondary imaging based on the array flow pattern of each array element at each region spectrum characteristic to obtain a characteristic-enhanced sound source image.
In step S9.1, the calculation formula of the array flow pattern of each array element at each region spectrum feature is as follows:
Figure 2016DEST_PATH_IMAGE062
，
wherein,
Figure DEST_PATH_IMAGE063
denotes the first
Figure 962013DEST_PATH_IMAGE035
The regional spectral characteristics of the individual sound-emanating regions,
Figure 521170DEST_PATH_IMAGE064
is shown as
Figure 30649DEST_PATH_IMAGE007
Array element is in
Figure 764862DEST_PATH_IMAGE035
The array flow pattern at the spectral features of the individual sound emanating areas,
Figure DEST_PATH_IMAGE065
the position of the sound source is represented,
Figure 727002DEST_PATH_IMAGE038
the number of the imaginary numbers is represented,
Figure 824271DEST_PATH_IMAGE039
which is indicative of the speed of sound,
Figure 142251DEST_PATH_IMAGE040
is shown as
Figure 299563DEST_PATH_IMAGE007
Coordinates of individual array elements.
In step S9.2, the calculation formula of the feature-enhanced sound source image is:
Figure 686682DEST_PATH_IMAGE066
，
wherein,
Figure 587641DEST_PATH_IMAGE042
representing a feature-enhanced sound source image,
Figure 25707DEST_PATH_IMAGE043
denotes the first
Figure 353920DEST_PATH_IMAGE044
The regional spectral characteristics of the individual sound-emanating regions,
Figure 290652DEST_PATH_IMAGE045
is shown as
Figure 214877DEST_PATH_IMAGE007
The area spectrum characteristics in the sound source signal received by each array element are
Figure 756717DEST_PATH_IMAGE046
The composition of (1).
The present embodiment can enhance a sound source image of a non-point sound source.
The above-mentioned embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements of the technical solutions of the present invention by those skilled in the art should fall within the protection scope of the present invention without departing from the design spirit of the present invention.

Claims (10)

1. A feature enhancement method for a non-point sound source image, comprising the steps of:
s1, determining the sound source position of a non-point sound source;
s2, pointing the microphone array of the acoustic imaging instrument to the sound source position to obtain an array focusing sound source signal;
s3, carrying out frequency spectrum search on the array focusing sound source signals to obtain a plurality of sound source frequency spectrum characteristics;
s4, performing band-pass filtering processing on the sound source signals received by the array elements according to a plurality of sound source frequency spectrum characteristics to obtain filtering signals received by the array elements, and performing frequency domain processing on the filtering signals to obtain frequency domain signals;
s5, according to the frequency domain signals received by all array elements at each moment
Figure QLYQS_2
Forming a time signal point corresponding to each time in the dimensional space,
Figure QLYQS_7
time signal point corresponding to time
Figure QLYQS_9
Has the coordinates of
Figure QLYQS_3
，
Figure QLYQS_5
To represent
Figure QLYQS_8
At a time, first
Figure QLYQS_11
The frequency domain signal received by an array element, i.e. at
Figure QLYQS_1
In a dimensional space of
Figure QLYQS_6
The coordinates of the individual dimensions are such that,
Figure QLYQS_10
representing the total number of microphone array elements,
Figure QLYQS_12
，
Figure QLYQS_4
represents the final time;
s6, clustering the signal points at all the moments based on a clustering algorithm to form
Figure QLYQS_13
Of individual, i.e. non-point, sound sources
Figure QLYQS_14
A sound emitting area;
s7, fitting the time signal points in each sound-emitting area to obtain each sound-emitting area
Figure QLYQS_15
The slope of the fitting straight line in different dimensions represents the weight coefficient of each array element in the sound production area corresponding to the fitting straight line;
s8, calculating to obtain sound source signals of all sounding areas at the final moment based on the filtering signals received by all array elements at the final moment and the weight coefficients of the signals received by all array elements at all sounding areas, and obtaining the area spectrum characteristics of all sounding areas based on the sound source signals of all sounding areas;
and S9, performing secondary imaging based on the frequency spectrum characteristics of each region, the positions of the array elements and the sound source position of the non-point sound source to obtain a sound source image with enhanced characteristics.
2. The method for enhancing characteristics of a non-point sound source image according to claim 1, wherein step S3 comprises the following steps:
s3.1, carrying out differential operation on the array focusing sound source signals:
Figure QLYQS_16
，
wherein,
Figure QLYQS_17
representing the differentiated array focused sound source signal,
Figure QLYQS_18
representing the array focused sound source signal,
Figure QLYQS_19
means differentiation;
s3.2, carrying out frequency spectrum search on the differentiated array focusing sound source signals to obtain a plurality of sound source frequency spectrum characteristics
Figure QLYQS_20
And each sound source frequency spectrum characteristic meets the following conditions:
Figure QLYQS_21
，
Figure QLYQS_22
，
wherein,
Figure QLYQS_23
is shown as
Figure QLYQS_24
The frequency spectrum characteristics of each sound source are determined,
Figure QLYQS_25
representing an imaging threshold.
3. The method for enhancing the characteristics of the sound source image of the non-point sound source according to claim 1, wherein between the step S5 and the step S6, the method further comprises the steps of:
normalizing the signal points at each moment to obtain a normalized signal point corresponding to each moment, wherein the normalized calculation formula is as follows:
Figure QLYQS_26
，
Figure QLYQS_27
，
wherein,
Figure QLYQS_28
to represent
Figure QLYQS_29
The frequency domain signal received by the first array element at a time,
Figure QLYQS_30
to represent
Figure QLYQS_31
A standardized signal point corresponding to the moment;
in step S6, the standardized signal points corresponding to each time are clustered based on a clustering algorithm to form
Figure QLYQS_32
Of individual, i.e. non-point, sound sources
Figure QLYQS_33
And a sound emitting area.
4. The method of claim 3, wherein the step S6 comprises the steps of:
s6.1, initialization
Figure QLYQS_34
A cluster center;
s6.2, calculating the distance between each standardized signal point and each clustering center, and dividing each standardized signal point into the class to which the clustering center closest to the standardized signal point belongs;
s6.3, based on all the standardized signal points in each class, recalculating to obtain the iterated signal
Figure QLYQS_35
A cluster center;
s6.4, repeating the step S6.2 to the step S6.3 until the preset iteration times are reached to obtain the final product
Figure QLYQS_36
Of individual, i.e. non-point, sound sources
Figure QLYQS_37
And a sound emitting area.
5. The method according to claim 4, wherein in step S7, the slopes of different dimensions of each sound emitting region are represented by a fitted straight line formed in a dimensional space:
Figure QLYQS_38
wherein
Figure QLYQS_39
Is shown as
Figure QLYQS_40
The fitting straight line corresponding to each sound production area is
Figure QLYQS_41
Slope of one dimension, i.e. representing the second
Figure QLYQS_42
Array element is in
Figure QLYQS_43
The individual voicing regions receive the weighting coefficients of the signals.
6. The method as claimed in claim 1, wherein in step S7, the time signal points in each sounding region are fitted by using least square method to obtain the time signal points in each sounding region
Figure QLYQS_44
A fitted straight line formed in a dimensional space.
7. The method of claim 1, wherein in step S8, the sound source signal of each sound emission area at the final time is calculated based on the following formula:
Figure QLYQS_45
，
wherein,
Figure QLYQS_46
indicates the last moment
Figure QLYQS_47
The filtered signals received by the individual array elements,
Figure QLYQS_48
is shown as
Figure QLYQS_49
Array element is in
Figure QLYQS_50
A sound production area receives signalsThe weight coefficient of (a) is,
Figure QLYQS_51
indicates the first time at the final moment
Figure QLYQS_52
Sound source signals of the individual sound emitting areas.
8. The method of enhancing features of a non-point sound source image according to claim 1, wherein the step S9 comprises the steps of:
s9.1, calculating an array flow pattern of each array element at each area spectrum characteristic position based on the area spectrum characteristics of each sound production area of the non-point sound source, the position of each array element and the sound source position;
and S9.2, performing secondary imaging based on the array flow pattern of each array element at each region spectrum feature to obtain a feature-enhanced sound source image.
9. The method according to claim 8, wherein in step S9.1, the formula for calculating the array flow pattern of each array element at each region spectral feature is as follows:
Figure QLYQS_53
，
wherein,
Figure QLYQS_56
is shown as
Figure QLYQS_59
The regional spectral characteristics of the individual sound-emanating regions,
Figure QLYQS_62
is shown as
Figure QLYQS_55
Array element is in
Figure QLYQS_60
The array flow pattern at the spectral features of the individual sound emanating areas,
Figure QLYQS_61
a position of a sound source is indicated,
Figure QLYQS_63
the number of the imaginary numbers is represented by,
Figure QLYQS_54
which is indicative of the speed of sound,
Figure QLYQS_57
is shown as
Figure QLYQS_58
Coordinates of individual array elements.
10. The method according to claim 9, wherein in step S9.2, the calculation formula of the feature-enhanced sound source image is:
Figure QLYQS_64
，
wherein,
Figure QLYQS_65
representing a feature-enhanced sound source image,
Figure QLYQS_66
is shown as
Figure QLYQS_67
The regional spectral characteristics of the individual sound-emanating regions,
Figure QLYQS_68
is shown as
Figure QLYQS_69
The area spectrum characteristics in the sound source signal received by each array element are
Figure QLYQS_70
The composition of (1).
CN202211524057.2A 2022-12-01 2022-12-01 Feature enhancement method for non-point sound source image Active CN115575896B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211524057.2A CN115575896B (en) 2022-12-01 2022-12-01 Feature enhancement method for non-point sound source image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211524057.2A CN115575896B (en) 2022-12-01 2022-12-01 Feature enhancement method for non-point sound source image

Publications (2)

Publication Number Publication Date
CN115575896A CN115575896A (en) 2023-01-06
CN115575896B true CN115575896B (en) 2023-03-10

Family

ID=84590473

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211524057.2A Active CN115575896B (en) 2022-12-01 2022-12-01 Feature enhancement method for non-point sound source image

Country Status (1)

Country Link
CN (1) CN115575896B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014215385A (en) * 2013-04-24 2014-11-17 日本電信電話株式会社 Model estimation system, sound source separation system, model estimation method, sound source separation method, and program
WO2018045973A1 (en) * 2016-09-08 2018-03-15 南京阿凡达机器人科技有限公司 Sound source localization method for robot, and system
KR20200038688A (en) * 2018-10-04 2020-04-14 서희 Apparatus and method for providing sound source
CN113884986A (en) * 2021-12-03 2022-01-04 杭州兆华电子有限公司 Beam focusing enhanced strong impact signal space-time domain joint detection method and system
CN114175144A (en) * 2019-07-30 2022-03-11 杜比实验室特许公司 Data enhancement for each generation of training acoustic models

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6635903B2 (en) * 2016-10-14 2020-01-29 日本電信電話株式会社 Sound source position estimating apparatus, sound source position estimating method, and program
CN107680593A (en) * 2017-10-13 2018-02-09 歌尔股份有限公司 The sound enhancement method and device of a kind of smart machine

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014215385A (en) * 2013-04-24 2014-11-17 日本電信電話株式会社 Model estimation system, sound source separation system, model estimation method, sound source separation method, and program
WO2018045973A1 (en) * 2016-09-08 2018-03-15 南京阿凡达机器人科技有限公司 Sound source localization method for robot, and system
KR20200038688A (en) * 2018-10-04 2020-04-14 서희 Apparatus and method for providing sound source
CN114175144A (en) * 2019-07-30 2022-03-11 杜比实验室特许公司 Data enhancement for each generation of training acoustic models
CN113884986A (en) * 2021-12-03 2022-01-04 杭州兆华电子有限公司 Beam focusing enhanced strong impact signal space-time domain joint detection method and system

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
ENHANCED POWER-NORMALIZED FEATURES FOR MANDARIN ROBUST SPEECH RECOGNITION BASED ON A VOICED-UNVOICED-SILENCE DECISION;Ying-Wei Tan et al.;《2014 IEEE China Summit & International Conference on Signal and Information Processing (ChinaSIP)》;20140613;第222-225页 *
基于频域ＩＣＡ的语音特征增强;吕钊 等;《振动与冲击》;20110225;第30卷(第2期);第238-257页 *

Also Published As

Publication number Publication date
CN115575896A (en) 2023-01-06

Similar Documents

Publication Publication Date Title
EP2746737B1 (en) Acoustic sensor apparatus and acoustic camera using a mems microphone array
JP5701164B2 (en) Position detection apparatus and position detection method
Ginn et al. Noise source identification techniques: simple to advanced applications
CN110491403A (en) Processing method, device, medium and the speech enabled equipment of audio signal
CA2496785A1 (en) Sound source search system
CN113868583B (en) Method and system for calculating sound source distance focused by subarray wave beams
WO2009145310A1 (en) Sound source separation and display method, and system thereof
JPH09512676A (en) Adaptive beamforming method and apparatus
CN110444220B (en) Multi-mode remote voice perception method and device
CN109683134A (en) A kind of high-resolution localization method towards rotation sound source
CN115435891A (en) Road vehicle sound power monitoring system based on vector microphone
Jing et al. Sound source localisation using a single acoustic vector sensor and multichannel microphone phased arrays
CN115575896B (en) Feature enhancement method for non-point sound source image
Prezelj et al. A novel approach to localization of environmental noise sources: Sub-windowing for time domain beamforming
CN114355290A (en) Sound source three-dimensional imaging method and system based on stereo array
CN114001816A (en) Acoustic imager audio acquisition system based on MPSOC
CN109061558A (en) A kind of sound collision detection and sound localization method based on deep learning
CN109254265A (en) A kind of whistle vehicle positioning method based on microphone array
CN116559778B (en) Vehicle whistle positioning method and system based on deep learning
Chen et al. Insight into split beam cross-correlator detector with the prewhitening technique
CN115061089B (en) Sound source positioning method, system, medium, equipment and device
CN112857560B (en) Acoustic imaging method based on sound frequency
CN116309921A (en) Delay summation acoustic imaging parallel acceleration method based on CUDA technology
Kerstens et al. An optimized planar MIMO array approach to in-air synthetic aperture sonar
Bianchi et al. High resolution imaging of acoustic reflections with spherical microphone arrays

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