CN105403860B - A kind of how sparse sound localization method related based on domination - Google Patents
A kind of how sparse sound localization method related based on domination Download PDFInfo
- Publication number
- CN105403860B CN105403860B CN201410451825.5A CN201410451825A CN105403860B CN 105403860 B CN105403860 B CN 105403860B CN 201410451825 A CN201410451825 A CN 201410451825A CN 105403860 B CN105403860 B CN 105403860B
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- frequency
- msubsup
- sound source
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Circuit For Audible Band Transducer (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
Abstract
The present invention relates to a kind of based on the how sparse sound localization method for dominating correlation, including:The sound-source signal received by microphone array is changed into digital audio signal;Extract the frequency spectrum of the digital audio signal of each microphone;The spatial correlation matrix on each frequency is calculated using the frequency spectrum of the digital audio signal of all microphones on the identical frequency of adjacent time;Extract the main characteristic vector of spatial correlation matrix;It is determined that on each frequency all microphones pair time delay set;Using the method for iteration, the azimuth for the sound source incident direction being top dog on each frequency is calculated;Statistical analysis is carried out to the azimuth of the sound source incident direction that is top dog on all frequencies, it is determined that final sound source incident direction and sound source number.This method considers acoustics robustness, it is adaptable to the real-time positioning of how sparse sound source.
Description
Technical field
The present invention relates to field of sound source location, more particularly to a kind of how sparse sound localization method related based on domination.
Background technology
Auditory localization includes simple sund source positioning and many auditory localizations, and auditory localization technology can be indicated where acoustic target
Dimensional orientation, important spatial information is provided for follow-up information gathering with processing.
Many auditory localization technologies occupy critical positions in the application of microphone array, and it can be used for teleconference, be wheat
Gram wind array indicates the direction that wave beam is focused on, and directional information is provided for meeting camera.
In field of sound source location, classified based on source signal and the auditory localization algorithm of spatial beams scanning is using most
For extensive algorithm, it is famous with the robustness under the conditions of noise and reverberation.However, there is a common spy in these algorithms
Point:Their cost function non-convex is non-recessed, causes the presence of the optimal estimation that multiple extreme values correspond to sound source position.Therefore, it is this kind of
Method generally seeks the globally optimal solution of sound source position by the way of lattice point traversal, causes amount of calculation to steeply rise, it is difficult to full
The demand positioned when full.
Above-mentioned problem is overcome based on the openness many sound localization methods of time-frequency domain, the method is assumed on each frequency only
There is a sound source to occupy ascendancy, the interference of other sound sources can be ignored.This hypothesis is by auditory localization more than one
Problem reduction is the simple sund source orientation problem on single frequency point, although considerably reduces amount of calculation, but is a lack of acoustics robust
Property causes this kind of method to be difficult to show excellent performance under complicated acoustic enviroment.
The content of the invention
It is computationally intensive present in existing many sound localization methods it is an object of the invention to overcome, lack acoustics robustness
Etc. defect, using sparse sound source time-frequency domain domination correlation, so as to provide a kind of robust and efficient many auditory localization sides
Method.
To achieve these goals, the present invention proposes a kind of how sparse sound localization method related based on domination, including:
Step 101), the sound-source signal received by microphone array is converted into digital audio signal, the microphone
Array includes K microphone;
Step 102), extract the frequency spectrum of the digital audio signal of each microphone;
Step 103), utilize the spectrometer of the digital audio signal of all microphones on the identical frequency of t adjacent time
Calculate the spatial correlation matrix on each frequency;
Step 104), the spatial correlation matrix on each frequency of t is decomposed, obtained on each frequency of t
Main characteristic vector, the main characteristic vector each component correspondence one microphone collection signal;
Step 105), M is asked on each frequency of t to microphone using the main characteristic vector on each frequency of t
Time delay set, the M be equal to K (K-1) 2;
Step 106), according to time delay set of the M to microphone on each frequency of t, using the method for iteration, meter
Calculate the azimuth for the sound source incident direction being top dog on each frequency of t;
Step 107), the azimuth of the sound source incident direction being top dog of all frequencies of t is counted
Analysis;
Step 108), export the t sound source incident direction and sound source number finally determined.
In above-mentioned technical proposal, in step 105) in, the main characteristic vector on each frequency of utilization t asks for t
Time delay set of the M to microphone includes on moment each frequency:
On t f frequencies, main characteristic vector is expressed as:[u1,t,f,u2,t,f,…uK,t,f], by p-th and q-th
The m (m=1,2 ..., M) of microphone composition is to the time phase difference of microphoneFor:
The operation of complex phase is asked in wherein ∠ () expressions,Represent angular frequency;
On t f frequencies, according to m to microphone apart from rmConstraint, obtains time phase aliasing set
Lm,t,f:
Wherein, c is the velocity of sound;
On t f frequencies, m is combined into the time delay collection of microphone:
Wherein,, representTime cycle in frequency.
In above-mentioned technical proposal, the step 106) further comprise:
Step 106-1), choose initial sound source incident direction
Step 106-2), a time-delay value is chosen from each time delay set;
OrderFrom each time delay set Bm,t,fOne time-delay value τ of middle selectionm,t,f, meet:
Wherein, gm=(gmx,gmy,gmz) represent direction unit vectors of the m to microphone line;
Step 106-3), ask for new weight coefficient wm,t,f;
Wherein:
δm,t,f=arccos (pTgm)-arccos(cτm,t,f/rm), m=1,2 ..., M
Step 106-4), calculate new sound source incident direction
Wherein:
Step 106-5), judgeWhether restrain;If a determination be made that certainly, it is transferred to step 106-6);Otherwise,
It is transferred to step 106-2) continue executing with;
Step 106-6), calculate the sound source incident direction being top dogOrientation
Angle.
In above-mentioned technical proposal, in step 106-1) in, it is described to choose initial sound source incident directionIncluding:
At the 360 ° × elevation angle in azimuth, 90 ° of lattice point spatially, uniformly chooses several sound source incident directions as candidate
Value, gathers { ψ labeled as vector1,ψ2,…ψH, H > 8 and for integer;For the sound source incident direction ψ of each candidateh, h=1,
2 ... H, calculate it and arrive the distance of all time delay set and be expressed as:
Wherein, % represents that floating-point remainder is operated;
On t f frequencies, initial sound source incident directionMeet:
In above-mentioned technical proposal, in step 107) in, the statistical analysis includes:Histogram analysis, clustering.
The advantage of the invention is that:
1st, by the domination similitude between adjacent time frequency, the information on adjacent time frequency is fully excavated, realization can
The estimated spatial position leaned on;
2nd, a kind of pairing time delay abstracting method of spatial correlation matrix based on time frequency block is proposed, this method is utilized
The weight coefficient of signal enhancing and time delay weighs reliability, it is achieved thereby that the how sparse sound localization method of robust.
Brief description of the drawings
Fig. 1 is a kind of flow chart of how sparse sound localization method based on ascendancy of the present invention;
Fig. 2 is the azimuthal method flow of sound source incident direction being top dog on each frequency of calculating of the present invention
Figure;
Fig. 3 is that the present invention enters column hisgram point to the sound source incident direction azimuth that is top dog on all frequencies
The flow chart of analysis;
Fig. 4 is the schematic diagram of the azimuth histogram analysis of the present invention.
Embodiment
Below by drawings and examples, technical scheme is described in further detail.
With reference to Fig. 1, method of the invention comprises the following steps:
Step 101), the sound-source signal received by microphone array is converted into digital audio signal;
The microphone array includes K microphone;
Step 102), digital audio signal is pre-processed, each Mike is extracted by Fast Fourier Transform (FFT) (FFT)
The frequency spectrum of the digital audio signal of wind;
It is described that digital audio signal progress pretreatment is included:To the first zero padding of the digital audio signal of each frame to N points, N
=2i, i is integer, and i >=8;Then, adding window is carried out to the digital audio signal of each frame or preemphasis is handled, windowed function
Using the peaceful window (hanning) of Hamming window (hamming) or Kazakhstan;
Fast Fourier Transform (FFT) is carried out to the digital audio signal of t, obtain t digital audio signal it is discrete
Frequency spectrum is
Wherein, yk,t,nRepresent that k-th of microphone of t gathers n-th of sampled point of signal, Yk,t,f(k=1,2 ... K, f
=0,1 ... N-1) represent that k-th of microphone of t gathers the Fourier Transform Coefficients of f-th of frequency of signal.
Step 103), when calculating t using the frequency spectrum of the digital audio signal of all microphones on the identical frequency of adjacent time
Carve the spatial correlation matrix of each frequency;
If xt,fFor t, one generated on f-th of frequency is complex vector located:xt,f={ Y1,t,f,Y2,t,f…YK,t,f, its
Autocorrelation matrix is:Wherein:()HRepresent conjugate transposition;
Plural autocorrelation matrix Rt,fIt is expressed as the average of autocorrelation matrix on the adjacent time frequency centered on f frequencies:
Wherein, A represents the frame number with t adjacent time;
xt,fSpatial correlation matrix be Rt,f。
Step 104), the spatial correlation matrix on each frequency of t is decomposed, obtained on each frequency of t
Main characteristic vector;The collection signal of each component one microphone of correspondence of vector;
Step 105), M is asked on each frequency of t to microphone using the main characteristic vector on each frequency of t
Time delay set, the M be equal to K (K-1) 2;Detailed process is:
On t f frequencies, main characteristic vector is expressed as:[u1,t,f,u2,t,f,…uK,t,f], by p-th and q-th
The m (m=1,2 ..., M) of microphone composition is to the time phase difference of microphoneFor:
The operation of complex phase is asked in wherein ∠ () expressions,Represent angular frequency;
On t f frequencies, according to m to microphone apart from rmConstraint, obtains time phase aliasing set
Lm,t,f:
Wherein, c is the velocity of sound;
On t f frequencies, m is combined into the time delay collection of microphone:
Wherein,, representTime cycle in frequency.
Step 106), according to time delay set of the M to microphone on each frequency of t, using the method for iteration, meter
Calculate the azimuth for the sound source incident direction being top dog on each frequency of t;
With reference to Fig. 2, specific implementation step is as follows:
Step 106-1), choose initial sound source incident direction
At the 360 ° × elevation angle in azimuth, 90 ° of lattice point spatially, uniformly chooses several sound source incident directions as candidate
Value, gathers { ψ labeled as vector1,ψ2,…ψH, H > 8 and for integer;For the sound source incident direction ψ of each candidateh, h=1,
2 ... H, calculate it and arrive the distance of all time delay set and be expressed as:
Wherein, gm=(gmx,gmy,gmz) direction unit vectors of the m to microphone line is represented, % represents floating-point remainder
Operation.
On t f frequencies, initial sound source incident directionMeet:
Step 106-2), a time-delay value is chosen from each time delay set;
OrderFrom each time delay set Bm,t,fOne time-delay value τ of middle selectionm,t,f, meet:
Step 106-3), ask for new weight coefficient wm,t,f;
Wherein:
δm,t,f=arccos (pTgm)-arccos(cτm,t,f/rm), m=1,2 ..., M
Step 106-4), calculate new sound source incident direction
Wherein:
Step 106-5), judgeWhether restrain;If a determination be made that certainly, it is transferred to step 106-6);Otherwise,
It is transferred to step 106-2) continue executing with;
In the present embodiment, judgeWhether convergent method is:
JudgeWhether threshold value ε is less than, wherein taking ε=0.01.
Step 106-6), calculate the sound source incident direction being top dogOrientation
Angle;Computational methods are:
Step 107) statistical is carried out to the azimuth of the sound source incident directions being top dog of all frequencies of t
Analysis;
The statistical analysis includes:Histogram analysis and clustering.
In the present embodiment, the azimuth of the sound source incident direction being top dog of all frequencies of t is carried out
Histogram analysis, with reference to Fig. 3, are comprised the following steps that:
Step 107-1), build and smooth azimuth histogram;
360 degree of azimuth is divided into 360 equal portions, histogram is built on these equal portions, and it is smooth using window function
Histogram.
Step 107-2), calculate histogram threshold value;
Histogram threshold value=average histogram value+(maximum Nogata value-average histogram value) × empirical coefficient;
Wherein empirical coefficient span is between 0 to 1;In the present embodiment, empirical coefficient value is 0.3.
Step 107-3), it is determined that final azimuth and azimuth number;
With reference to Fig. 4, choose peak value in histogram and be more than the peak of histogram threshold value as final azimuth, of peak value
Number is azimuthal number.
Step 107-4), it is determined that final sound source incident direction and sound source number.
The corresponding sound source incident direction in this azimuth is obtained according to the azimuth of selection, azimuthal number is that sound source is incident
The number in direction.
Step 108), export the t sound source incident direction and sound source number finally determined.
In other embodiments, can be to the orientation of the sound source incident direction being top dog of all frequencies of t
Angle carries out clustering, and specific processing method belongs to common knowledge, will not be described here.
Claims (5)
1. it is a kind of based on the how sparse sound localization method for dominating correlation, including:
Step 101), the sound-source signal received by microphone array is converted into digital audio signal, the microphone array
Including K microphone;
Step 102), extract the frequency spectrum of the digital audio signal of each microphone;
Step 103), calculate every using the frequency spectrum of the digital audio signal of all microphones on the identical frequency of t adjacent time
Spatial correlation matrix on individual frequency;
Step 104), the spatial correlation matrix on each frequency of t is decomposed, the master on each frequency of t is obtained
Characteristic vector, the collection signal of each component one microphone of correspondence of the main characteristic vector;
Step 105), using the main characteristic vector on each frequency of t ask for M on each frequency of t to microphone when
Between postpone set, the M be equal to K (K-1)/2;
Step 106), according to time delay set of the M to microphone on each frequency of t, using the method for iteration, calculate t
The azimuth for the sound source incident direction being top dog on moment each frequency;
Step 107), statistical analysis is carried out to the azimuth of the sound source incident direction being top dog of all frequencies of t;
Step 108), export the t sound source incident direction and sound source number finally determined.
2. it is according to claim 1 based on the how sparse sound localization method for dominating correlation, it is characterised in that in step
105) in, the main characteristic vector on each frequency of utilization t asks for times of the M to microphone on each frequency of t
Delay set includes:
On t f frequencies, main characteristic vector is expressed as:[u1,t,f,u2,t,f,…uK,t,f], by p-th and q-th of Mike
The m (m=1,2 ..., M) of wind composition is to the time phase difference of microphoneFor:
The operation of complex phase is asked in wherein ∠ () expressions,Represent angular frequency;
On t f frequencies, according to m to microphone apart from rmConstraint, obtains time phase aliasing set Lm,t,f:
Wherein, c is the velocity of sound;
On t f frequencies, m is combined into the time delay collection of microphone:
<mrow>
<msub>
<mi>B</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<mo>=</mo>
<mo>{</mo>
<msub>
<mover>
<mi>&tau;</mi>
<mo>^</mo>
</mover>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<mo>=</mo>
<msub>
<mover>
<mi>&tau;</mi>
<mo>&OverBar;</mo>
</mover>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>l</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<msub>
<mi>T</mi>
<mi>f</mi>
</msub>
<mo>|</mo>
<msub>
<mi>l</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<mo>&Element;</mo>
<msub>
<mi>L</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<mo>}</mo>
<mo>,</mo>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<mi>M</mi>
</mrow>
Wherein,RepresentTime cycle in frequency.
3. it is according to claim 2 based on the how sparse sound localization method for dominating correlation, it is characterised in that step 106)
Further comprise:
Step 106-1), choose initial sound source incident directionOrder
Step 106-2), a time-delay value is chosen from each time delay set;
OrderFrom each time delay set Bm,t,fOne time-delay value τ of middle selectionm,t,f, meet:
<mrow>
<msub>
<mi>&tau;</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<mo>=</mo>
<mi>arg</mi>
<munder>
<mi>min</mi>
<mrow>
<mi>&tau;</mi>
<mo>&Element;</mo>
<msub>
<mi>B</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
</mrow>
</munder>
<mrow>
<mo>|</mo>
<mrow>
<msubsup>
<mi>pg</mi>
<mi>m</mi>
<mi>T</mi>
</msubsup>
<msub>
<mi>r</mi>
<mi>m</mi>
</msub>
<mo>-</mo>
<mi>c</mi>
<mi>&tau;</mi>
</mrow>
<mo>|</mo>
</mrow>
<mo>,</mo>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<mi>M</mi>
<mo>;</mo>
</mrow>
1
Wherein, gm=(gmx,gmy,gmz) represent direction unit vectors of the m to microphone line;
Step 106-3), ask for new weight coefficient wm,t,f;
<mrow>
<msub>
<mi>w</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<mo>=</mo>
<mi>exp</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>&delta;</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
<mn>2</mn>
</msubsup>
<mo>/</mo>
<msup>
<mi>&sigma;</mi>
<mn>2</mn>
</msup>
<mo>)</mo>
</mrow>
<mo>/</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>M</mi>
</munderover>
<mi>exp</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>&delta;</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
<mn>2</mn>
</msubsup>
<mo>/</mo>
<msup>
<mi>&sigma;</mi>
<mn>2</mn>
</msup>
<mo>)</mo>
</mrow>
<mo>,</mo>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<mi>M</mi>
</mrow>
Wherein:
δm,t,f=arccos (pTgm)-arccos(cτm,t,f/rm), m=1,2 ..., M
<mrow>
<msup>
<mi>&sigma;</mi>
<mn>2</mn>
</msup>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>M</mi>
</munderover>
<msubsup>
<mi>&delta;</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
<mn>2</mn>
</msubsup>
<mo>/</mo>
<mi>M</mi>
</mrow>
Step 106-4), calculate new sound source incident direction
<mrow>
<mfenced open = "(" close = ")">
<mtable>
<mtr>
<mtd>
<msub>
<mover>
<mi>&gamma;</mi>
<mo>^</mo>
</mover>
<mrow>
<mn>1</mn>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<msub>
<mover>
<mi>&gamma;</mi>
<mo>^</mo>
</mover>
<mrow>
<mn>2</mn>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>=</mo>
<msup>
<mrow>
<mo>&lsqb;</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>M</mi>
</munderover>
<msub>
<mi>w</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<msubsup>
<mi>g</mi>
<mi>m</mi>
<mo>&prime;</mo>
</msubsup>
<msubsup>
<mi>g</mi>
<mi>m</mi>
<mrow>
<mo>&prime;</mo>
<mi>T</mi>
</mrow>
</msubsup>
<mo>&rsqb;</mo>
</mrow>
<mrow>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msup>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>M</mi>
</munderover>
<msub>
<mi>cw</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<msub>
<mi>&tau;</mi>
<mrow>
<mi>m</mi>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<msubsup>
<mi>g</mi>
<mi>m</mi>
<mrow>
<mo>&prime;</mo>
<mi>T</mi>
</mrow>
</msubsup>
<mo>/</mo>
<msub>
<mi>r</mi>
<mi>m</mi>
</msub>
</mrow>
<mrow>
<msub>
<mover>
<mi>&gamma;</mi>
<mo>^</mo>
</mover>
<mrow>
<mn>3</mn>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
</msub>
<mo>=</mo>
<msqrt>
<mrow>
<mn>1</mn>
<mo>-</mo>
<msubsup>
<mover>
<mi>&gamma;</mi>
<mo>^</mo>
</mover>
<mrow>
<mn>1</mn>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
<mn>2</mn>
</msubsup>
<mo>-</mo>
<msubsup>
<mover>
<mi>&gamma;</mi>
<mo>^</mo>
</mover>
<mrow>
<mn>2</mn>
<mo>,</mo>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
</mrow>
<mn>2</mn>
</msubsup>
</mrow>
</msqrt>
</mrow>
Wherein:g′m=(gmx,gmy);
Step 106-5), judgeWhether restrain;If a determination be made that certainly, it is transferred to step 106-6);Otherwise, makeIt is transferred to step 106-2) continue executing with;
Step 106-6), calculate the sound source incident direction being top dogAzimuth.
4. it is according to claim 3 based on the how sparse sound localization method for dominating correlation, it is characterised in that in step
It is described to choose initial sound source incident direction in 106-1)Including:
At the 360 ° × elevation angle in azimuth, 90 ° of lattice point spatially, uniformly chooses several sound source incident directions as candidate value, mark
It is designated as vector set { ψ1,ψ2,…ψH, H > 8 and for integer;For the sound source incident direction ψ of each candidateh, h=1,2 ... H,
It is calculated to arrive the distance of all time delay set and be expressed as:
<mrow>
<msub>
<mi>b</mi>
<mi>f</mi>
</msub>
<mrow>
<mo>(</mo>
<msub>
<mi>&psi;</mi>
<mi>h</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>m</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>M</mi>
</munderover>
<mo>&lsqb;</mo>
<msubsup>
<mi>&psi;</mi>
<mi>H</mi>
<mi>T</mi>
</msubsup>
<msub>
<mi>g</mi>
<mi>m</mi>
</msub>
<msub>
<mi>r</mi>
<mi>m</mi>
</msub>
<mo>-</mo>
<mi>c</mi>
<mi>&tau;</mi>
<mo>)</mo>
<mi>%</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>cT</mi>
<mi>f</mi>
</msub>
<mo>)</mo>
</mrow>
<msup>
<mo>&rsqb;</mo>
<mn>2</mn>
</msup>
</mrow>
Wherein, % represents that floating-point remainder is operated;
On t f frequencies, initial sound source incident directionMeet:
<mrow>
<msub>
<mover>
<mi>&gamma;</mi>
<mo>^</mo>
</mover>
<mrow>
<mi>t</mi>
<mo>,</mo>
<mi>f</mi>
<mo>,</mo>
<mn>0</mn>
</mrow>
</msub>
<mo>=</mo>
<munder>
<mrow>
<mi>arg</mi>
<mi>min</mi>
</mrow>
<mrow>
<mn>1</mn>
<mo>&le;</mo>
<mi>h</mi>
<mo>&le;</mo>
<mi>H</mi>
</mrow>
</munder>
<msub>
<mi>b</mi>
<mi>f</mi>
</msub>
<mrow>
<mo>(</mo>
<msub>
<mi>&psi;</mi>
<mi>h</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>.</mo>
</mrow>
5. it is according to claim 1 based on the how sparse sound localization method for dominating correlation, it is characterised in that in step
107) in, the statistical analysis includes:Histogram analysis, clustering.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410451825.5A CN105403860B (en) | 2014-08-19 | 2014-08-19 | A kind of how sparse sound localization method related based on domination |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410451825.5A CN105403860B (en) | 2014-08-19 | 2014-08-19 | A kind of how sparse sound localization method related based on domination |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105403860A CN105403860A (en) | 2016-03-16 |
CN105403860B true CN105403860B (en) | 2017-10-31 |
Family
ID=55469461
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410451825.5A Active CN105403860B (en) | 2014-08-19 | 2014-08-19 | A kind of how sparse sound localization method related based on domination |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105403860B (en) |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106054196B (en) * | 2016-05-20 | 2018-03-16 | 中国民用航空总局第二研究所 | The acoustics localization method and device of a kind of airdrome scene target |
CN106019229B (en) * | 2016-05-20 | 2018-03-16 | 中国民用航空总局第二研究所 | Airdrome scene target acoustical localization method, sound sensing device and system |
CN108269583B (en) * | 2017-01-03 | 2021-07-30 | 中国科学院声学研究所 | Voice separation method based on time delay histogram |
CN108398664B (en) * | 2017-02-07 | 2020-09-08 | 中国科学院声学研究所 | Analytic spatial de-aliasing method for microphone array |
CN109920444B (en) * | 2017-12-13 | 2021-04-27 | 中国电信股份有限公司 | Echo time delay detection method and device and computer readable storage medium |
CN108152788A (en) * | 2017-12-22 | 2018-06-12 | 西安Tcl软件开发有限公司 | Sound-source follow-up method, sound-source follow-up equipment and computer readable storage medium |
CN109975762B (en) * | 2017-12-28 | 2021-05-18 | 中国科学院声学研究所 | Underwater sound source positioning method |
CN110133595B (en) * | 2018-02-09 | 2023-05-23 | 北京搜狗科技发展有限公司 | Sound source direction finding method and device for sound source direction finding |
CN109001681A (en) * | 2018-06-25 | 2018-12-14 | 大连大学 | The method of construction compression observing matrix in more auditory localizations |
CN109001680A (en) * | 2018-06-25 | 2018-12-14 | 大连大学 | The sparse optimization algorithm of block in auditory localization |
CN110146895B (en) | 2019-05-16 | 2021-04-20 | 浙江大学 | Acoustic velocity profile inversion method based on inverted multi-beam echometer |
CN110398716B (en) * | 2019-08-23 | 2021-05-28 | 北京工业大学 | Multi-sound-source positioning method utilizing sparse component equalization among sound sources |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102103200A (en) * | 2010-11-29 | 2011-06-22 | 清华大学 | Acoustic source spatial positioning method for distributed asynchronous acoustic sensor |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030072456A1 (en) * | 2001-10-17 | 2003-04-17 | David Graumann | Acoustic source localization by phase signature |
-
2014
- 2014-08-19 CN CN201410451825.5A patent/CN105403860B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102103200A (en) * | 2010-11-29 | 2011-06-22 | 清华大学 | Acoustic source spatial positioning method for distributed asynchronous acoustic sensor |
Non-Patent Citations (2)
Title |
---|
A Novel Multiple Sparse Source Localization Using Triangular Pyramid Microphone Array;Mengqi Ren等;《IEEE SIGNAL PROCESSING LETTERS》;20120229;第19卷(第2期);全文 * |
几种基本时间延迟估计方法及其相互关系;邱天爽 等;《Journal of Dalian University of Technology》;19960731;全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN105403860A (en) | 2016-03-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105403860B (en) | A kind of how sparse sound localization method related based on domination | |
Cao et al. | Polyphonic sound event detection and localization using a two-stage strategy | |
CN106405501B (en) | A kind of simple sund source localization method returned based on phase difference | |
CN109272989B (en) | Voice wake-up method, apparatus and computer readable storage medium | |
CN107102296B (en) | Sound source positioning system based on distributed microphone array | |
CN103076593B (en) | Sound source localization method and device | |
JP3906230B2 (en) | Acoustic signal processing apparatus, acoustic signal processing method, acoustic signal processing program, and computer-readable recording medium recording the acoustic signal processing program | |
US9264806B2 (en) | Apparatus and method for tracking locations of plurality of sound sources | |
CN111474521B (en) | Sound source positioning method based on microphone array in multipath environment | |
JP2021110938A5 (en) | ||
CN106093921A (en) | Acoustic vector sensor array broadband based on sparse resolution theory direction-finding method | |
Shimada et al. | Ensemble of ACCDOA-and EINV2-based systems with D3Nets and impulse response simulation for sound event localization and detection | |
Huleihel et al. | Spherical array processing for acoustic analysis using room impulse responses and time-domain smoothing | |
CN105976827A (en) | Integrated-learning-based indoor sound source positioning method | |
CN110610718A (en) | Method and device for extracting expected sound source voice signal | |
CN105607042A (en) | Method for locating sound source through microphone array time delay estimation | |
WO2016119388A1 (en) | Method and device for constructing focus covariance matrix on the basis of voice signal | |
JP2006340391A (en) | Acoustic signal processing apparatus, acoustic signal processing method, acoustic signal processing program, and computer readable record medium which records acoustic signal processing program | |
CN109541573A (en) | A kind of element position calibration method being bent hydrophone array | |
CN109901114B (en) | Time delay estimation method suitable for sound source positioning | |
CN111352075B (en) | Underwater multi-sound-source positioning method and system based on deep learning | |
CN108398664B (en) | Analytic spatial de-aliasing method for microphone array | |
Yang et al. | Supervised direct-path relative transfer function learning for binaural sound source localization | |
JP2015037207A (en) | Sound field gathering/reproducing device, method and program | |
CN108269583B (en) | Voice separation method based on time delay histogram |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |