CN101499280B - Spacing parameter choosing method and apparatus based on spacing perception entropy judgement - Google Patents

Spacing parameter choosing method and apparatus based on spacing perception entropy judgement Download PDF

Info

Publication number
CN101499280B
CN101499280B CN2009100610285A CN200910061028A CN101499280B CN 101499280 B CN101499280 B CN 101499280B CN 2009100610285 A CN2009100610285 A CN 2009100610285A CN 200910061028 A CN200910061028 A CN 200910061028A CN 101499280 B CN101499280 B CN 101499280B
Authority
CN
China
Prior art keywords
spatial parameter
entropy
frequency band
spatial
parameter
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.)
Expired - Fee Related
Application number
CN2009100610285A
Other languages
Chinese (zh)
Other versions
CN101499280A (en
Inventor
胡瑞敏
陈水仙
陈琪
张磊
陈文琴
王恒
陈冰
曹晟
杨健仁
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.)
Wuhan University WHU
Original Assignee
Wuhan University WHU
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 Wuhan University WHU filed Critical Wuhan University WHU
Priority to CN2009100610285A priority Critical patent/CN101499280B/en
Publication of CN101499280A publication Critical patent/CN101499280A/en
Application granted granted Critical
Publication of CN101499280B publication Critical patent/CN101499280B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Compression, Expansion, Code Conversion, And Decoders (AREA)

Abstract

The invention proposes a spatial parameter selection method based on spatial awareness entropy judgment and a device thereof. The method comprises the steps of: pre-selecting spatial parameters by comparing the critical perception threshold values of spatial parameters of adjacent frequency bands and amending the pre-selecting result according to the spatial awareness entropy of an actual signal and the magnitude of the fluctuation thereof, thus realizing low-complexity and high-efficiency spatial parameter selection. The invention can optimize spatial parameter selection and realize high-efficiency spatial audio coding.

Description

Space parameter selection method and device thereof based on spacing perception entropy judgement
Technical field
The present invention relates to the spatial audio coding field, especially relate to space parameter selection method and device thereof based on spacing perception entropy judgement.
Background technology
The subjective attribute of sound belongs to the monaural attribute as intensity, tone, tone color etc.; The space attribute of sound is as belonging to the ears attribute to localization of sound (direction and distance) and to the acoustic space characteristic perception of environment etc.Human mainly being meant by the obtainable spatial information of ears audition comprises clues such as distance, horizontal angle of deviation and elevation angle at the locus of sound source.Space psychologic acoustics result of study shows, the direct sound wave that sound source produces arrives ears behind the diffraction of head etc., produce ears mistiming (Interaural Time Difference is hereinafter to be referred as ITD) and binaural sound strong difference (Interaural Level Difference is hereinafter to be referred as ILD).Auditory system is utilized the level orientation that ITD and ILD can localization of sound source.The degree of correlation between another spatial parameter ear (Interaural Correlation is hereinafter to be referred as IC) expression left and right sides ear entering tone similarity is in order to determine the sound field width of sound source.
In information theory, Shannon entropy is used as measuring of information.1988, Johnston proposes perceptual entropy (Perceptual Entropy is hereinafter to be referred as PE) on the basis of Shannon entropy, but measuring as the sensing audio signal message, that is to say, but how many bits of minimum needs could be represented a tonic train under the prerequisite of not having perceptual distortion.The sensing audio encoding system of existing main flow with PE as an important psychologic acoustics index, as MP3 and AAC scrambler.Spatial information is the spatial parameter of people's ear perception sound field effect, has the redundancy in the perception, remains the guidance of entropy it is removed.Yet its meaning has exceeded the category of perceptual entropy, needs a brand-new entropy badly and explains appreciable spatial information amount, is used in reference to the coding of guide space information.
Spacing perception entropy SPE (Spatial Perceptual Entropy is hereinafter to be referred as SPE) represents the size of the spatial information amount that can perceivedly arrive in the multi-channel audio signal.The proposition of spacing perception entropy is utilizing parameter to carry out on the basis of auditory localization, has solved the problem of the estimation perception information amount size that spatial parameter contained.It is based on spatial perception measure information model, and the combined input signal self character can be become the another effective tool of guiding space audio coding by the spatial information of people's ear identification in the quantitative description multi-channel audio signal.
Yet current research to space parameter selection method is guidance with the duplexing theory of classics usually, promptly since ITD and ILD when low frequency and high frequency, play a major role respectively.In the coded system practical application, extract ITD in frequency less than the frequency range of 1kHz, and extract ILD greater than the frequency range of 1.5kHz in frequency.This single space parameter selection method, each spatial parameter is isolated the extraction of coming, both ignored the critical apperceive characteristic of spatial parameter in the full range band, the difference of the space parameter space quantity of information in the actual signal of having no basis again, choose and carry the bigger spatial parameter of spatial information and carry out audio coding, there is the subjective redundancy between spatial parameter, limited the further raising of multi-channel audio coding efficient.
Summary of the invention
The present invention proposes space parameter selection method and device thereof based on spacing perception entropy judgement, purpose is the guiding space audio coding, effectively removes exist in the space parameter selection method subjective redundant.
The invention provides a kind of space parameter selection method, may further comprise the steps based on spacing perception entropy judgement:
Step 1 is carried out spatial parameter and is chosen in advance, comprises following three steps,
Step 1.1 according to the fluctuation situation of each spatial parameter critical perception threshold value in each frequency band, is carried out the separate bands grouping of single spatial parameter;
Step 1.2 is carried out the combine frequency band grouping of all spatial parameters on the separate bands packet by packet basis of single spatial parameter, whether each spatial parameter is chosen respectively in the combine frequency band grouping adjudicated in advance;
Step 1.3, comprehensive pre-judgement result to all spatial parameters obtains spatial parameter and chooses scheme in advance;
Step 2 is carried out time-frequency conversion and frequency band division to every frame input signal, extracts each spatial parameter, calculates the spacing perception entropy and the moving degree of entropy wave thereof of each spatial parameter;
Step 3 compares the spacing perception entropy of each spatial parameter in the step 2 respectively with its corresponding threshold, carry out the spatial parameter selection project setting according to comparative result at every frame input signal,
If spacing perception entropy is greater than 0 and at very big thresholding max αBelow, the moving degree of the spacing perception entropy of the different spaces parameter of more every frame input signal and entropy wave thereof if entropy wave moving degree maximum and spacing perception entropy are not minimum, all carries out choosing of this spatial parameter on the full range band; Otherwise the spatial parameter that keeps step 1 gained on the full range band is chosen scheme in advance;
If spacing perception entropy less than 0, does not carry out choosing of this spatial parameter on the full range band;
If spacing perception entropy is at very big thresholding max αMore than, all carry out choosing of this spatial parameter on the full range band.
And, in step 1.1, the separate bands of the carrying out single spatial parameter concrete mode of dividing into groups is, be analytic target at first with the critical perception threshold value of single spatial parameter in each frequency band, the critical perception threshold value that compares nearby frequency bands, the critical perception threshold value that numerical value is close is classified as one group, thereby finishes the separate bands grouping of single spatial parameter;
In step 1.2, the boundaries of packets of separate bands grouping of getting each spatial parameter is finished the combine frequency band grouping as the boundaries of packets of the combine frequency band grouping of all spatial parameters according to the boundaries of packets of combine frequency band grouping; Whether each spatial parameter is chosen the concrete mode of adjudicating in advance respectively in the combine frequency band grouping be, the critical perception threshold size of more single spatial parameter between group, spatial parameter to the less group of critical perception threshold value is chosen, otherwise will not choose.
And, in step 2, when calculating moving the spending of entropy wave of present frame, with reference to the spacing perception entropy of present frame and the spacing perception entropy of frame before.
And the space perceptual entropy is greater than 0 and at very big thresholding max in step 3 αWhen following, at first the moving degree of the entropy wave of comparison step 2 each spatial parameter of gained is chosen the maximum spatial parameter of the moving degree of entropy wave; The spacing perception entropy that compares each spatial parameter then, the spacing perception entropy of the spatial parameter maximum as if the moving degree of entropy wave is not a minimum, then all carries out choosing of this spatial parameter on the full range band, otherwise the spatial parameter of reservation step 1 gained is chosen scheme in advance on the full range band.
The present invention also provides corresponding spatial parameter selection device, includes the moving degree of spatial parameter preliminary election delivery piece, spacing perception entropy and entropy wave computing module, spacing perception entropy judgement module and spatial parameter selection adjusting module;
Described spatial parameter preliminary election delivery piece comprises separate bands grouped element, the combine frequency band grouped element that order links to each other and chooses pre-judgement unit;
The moving degree of described spacing perception entropy and entropy wave computing module comprises time-frequency conversion unit, frequency band division unit, spatial parameter extraction unit, frequency band SPE computing unit, divides frame SPE computing unit and fluctuation degree computing unit, time-frequency conversion unit, frequency band division unit, spatial parameter extraction unit and frequency band SPE computing unit are connected successively, and the output of frequency band SPE computing unit links to each other with fluctuation degree computing unit with minute frame SPE computing unit respectively;
Described spacing perception entropy judgement module comprises threshold value decision unit and fluctuation degree decision unit, divide the output of frame SPE computing unit to enter the threshold value decision unit, divide the output of frame SPE computing unit, fluctuation degree computing unit and threshold value decision unit to enter fluctuation degree decision unit;
The output of described spacing perception entropy judgement module and spatial parameter preliminary election delivery piece inserts the spatial parameter selection adjusting module.
The present invention is according to spatial parameter critical perception threshold value (Just Noticeable Difference in each frequency band, hereinafter to be referred as JND) difference, formulate the scheme of choosing in advance of spatial parameter, the size and the variation of spacing perception entropy in each frame signal of estimation spatial parameter, append to choose and carry the spatial parameter that more spatial information is but ignored by former choosing method, to such an extent as to or reject the spatial parameter that the spatial information amount seldom is not perceived by the human ear, optimize spatial parameter selection, realize high-efficiency spatial audio coding.
Description of drawings
Fig. 1 is a frequency band group technology synoptic diagram of the present invention;
Fig. 2 chooses process flow diagram in advance for spatial parameter of the present invention;
Fig. 3 is a method schematic diagram of the present invention;
Fig. 4 is the frequency band grouping implementation process synoptic diagram of the embodiment of the invention;
Fig. 5 is spacing perception entropy of the present invention and the moving degree of entropy wave calculation flow chart;
Fig. 6 is that the spatial parameter selection that the present invention is based on spacing perception entropy judgement is adjusted process flow diagram.
Embodiment
The technical scheme of space parameter selection method of the present invention may further comprise the steps:
Step 1 is carried out spatial parameter and is chosen in advance, may further comprise the steps,
Step 1.1 according to the fluctuation situation of each spatial parameter critical perception threshold value in each frequency band, is carried out the separate bands grouping of single spatial parameter;
Step 1.2 is carried out the combine frequency band grouping of all spatial parameters on the separate bands packet by packet basis of single spatial parameter, whether each spatial parameter is chosen respectively in the combine frequency band grouping adjudicated in advance;
Step 1.3, comprehensive pre-judgement result to all spatial parameters obtains spatial parameter and chooses scheme in advance.
The invention provides further spatial parameter and choose technical scheme in advance: be analytic target at first with the critical perception threshold value of single spatial parameter in each frequency band, the critical perception threshold value that compares nearby frequency bands, the critical perception threshold value that numerical value is close is classified as one group, thereby finishes the separate bands grouping of single spatial parameter; The boundaries of packets of separate bands grouping of getting each spatial parameter is finished the combine frequency band grouping as the boundaries of packets of the combine frequency band grouping of all spatial parameters according to the boundaries of packets of combine frequency band grouping.Referring to Fig. 1, boundaries of packets in the separate bands grouping of a spatial parameter is a frequency band and b frequency band, boundaries of packets in the separate bands grouping of another spatial parameter is the c frequency band, and the boundaries of packets of the combine frequency band of these two spatial parameters grouping is a frequency band, b frequency band and c frequency band so.During concrete enforcement,, adjacent packets border more approaching in the combine frequency band grouping can be merged in order to reduce code check or complexity.
After finishing the combine frequency band grouping, the critical perception threshold size of more single spatial parameter between group chosen the spatial parameter of the less group of critical perception threshold value, otherwise will not be chosen.Because the unit of the critical perception threshold value of different spaces parameter is often different, inconvenience directly quantizes comparison, can distinguish during concrete enforcement earlier by the critical perception threshold size of single spatial parameter between group, divide this spatial parameter and between group, choose priority, the critical perception threshold value is more little, and priority is high more; The priority of comprehensive then all spatial parameters, for each group is chosen the high spatial parameter of priority, the spatial parameter that constitutes on the full range band is chosen scheme in advance.Therefore might in certain grouping of full range band, there be an above spatial parameter to be selected.Concrete how according to the priority of critical perception threshold value division spatial parameter between group, can require to set according to concrete complexity and code check, many more complexities of the rank of division and code check are high more.
Step 2 is carried out time-frequency conversion and frequency band division to every frame input signal, extracts each spatial parameter, calculates the spacing perception entropy and the moving degree of entropy wave thereof of each spatial parameter.During concrete enforcement, the frequency band division method can be with division rule with reference to bark.
Spacing perception entropy can be asked for reference to definition, belongs to prior art, and adoptable concrete computation process is as follows when the invention provides enforcement:
(1) calculates in the i frame signal space parameter at the spacing perception entropy spe of frequency band b i(b), calculating formula is as follows:
spe i ( b ) = μ · log ( | α ( b ) Δα ( b ) | ) (formula 1)
Wherein, i is a frame number, and b is band designation, and α represents the spatial parameter chosen, and α (b) is the numerical value of spatial parameter α at frequency band b, and Δ α (b) is the JND of spatial parameter α at frequency band b, and μ is the amplitude compression of sense of hearing perception.
(2) the spacing perception entropy SPE of computer memory parameter i frame i, calculating formula is as follows:
SPE i = 1 N Σ b = 1 n spe i ( b ) (formula 2)
Wherein, b is band designation, spe i(b) be that the space parameter is at the spacing perception entropy of frequency band b in the i frame signal, N is the frame length of conversion, the frequency band number of n for dividing.
When calculating moving the spending of entropy wave of present frame, not only use the spacing perception entropy of present frame, and use the spacing perception entropy of frame before, can accurately reflect the fluctuation situation of spacing perception entropy like this.During concrete enforcement, can adopt multiple mode, for example by calculating the moving degree of entropy wave that every frame signal and the standard deviation of the spacing perception entropy of previous frame signal characterize present frame, perhaps the entropy wave that calculates present frame by the every frame signal and the standard deviation of the spacing perception entropy of all frame signals before moves and spends.
Step 3 compares the spacing perception entropy of each spatial parameter in the step 2 respectively with its corresponding threshold, carry out the spatial parameter selection project setting according to comparative result at every frame input signal,
If spacing perception entropy is greater than 0 and at very big thresholding max αBelow, the moving degree of the spacing perception entropy of the different spaces parameter of more every frame input signal and entropy wave thereof if entropy wave moving degree maximum and spacing perception entropy are not minimum, all carries out choosing of this spatial parameter on the full range band; Otherwise the spatial parameter that keeps step 1 gained on the full range band is chosen scheme in advance;
If spacing perception entropy less than 0, does not carry out choosing of this spatial parameter on the full range band;
If spacing perception entropy is at very big thresholding max αMore than, all carry out choosing of this spatial parameter on the full range band.
Adopting the threshold decision mode, is the spatial parameter that carries more spatial informations in order to select.When spacing perception entropy less than 0 the time, illustrate that the quantity of information that this spatial parameter can not be perceived by the human ear and carry ignores, itself is unimportant, can choose.The very big thresholding max that sets greater than value rule of thumb when spacing perception entropy αThe time, prove absolutely the importance of this spatial parameter, so on the full range band, all carried out choosing of this spatial parameter.
During concrete enforcement in step 3 the space perceptual entropy greater than 0 and at very big thresholding max αWhen following, processing can be chosen the maximum spatial parameter of the moving degree of entropy wave and begin from the moving degree of the entropy wave of comparison step 2 each spatial parameter of gained at first; The spacing perception entropy that compares each spatial parameter then.If the spacing perception entropy of the spatial parameter that the moving degree of entropy wave is maximum is not minimum, then choose scheme in advance at the spatial parameter of this spatial parameter set-up procedure 1; Otherwise the spatial parameter that keeps step 1 gained on the full range band is chosen scheme in advance, and this comprises that it is not that maximum and spacing perception entropy minimum, the moving degree of entropy wave are not that maximum and spacing perception entropy also are not two kinds of situations of minimum that entropy wave moves degree.
Space parameter selection method provided by the invention can adopt computer software technology to realize, also can adopt modularity design technology to realize that isolated plant carries out spatial parameter selection work.Device technique scheme provided by the invention is: include the moving degree of spatial parameter preliminary election delivery piece, spacing perception entropy and entropy wave computing module, spacing perception entropy judgement module and spatial parameter selection adjusting module;
Described spatial parameter preliminary election delivery piece comprises separate bands grouped element, the combine frequency band grouped element that order links to each other and chooses pre-judgement unit, be used for carrying out respectively the combine frequency band grouped element of separate bands grouping, all spatial parameters of single spatial parameter, whether each spatial parameter chosen respectively in the combine frequency band grouping adjudicated in advance, can be referring to Fig. 2;
The moving degree of described spacing perception entropy and entropy wave computing module comprises time-frequency conversion unit, frequency band division unit, spatial parameter extraction unit, frequency band SPE computing unit, divides frame SPE computing unit and fluctuation degree computing unit, time-frequency conversion unit, frequency band division unit, spatial parameter extraction unit and frequency band SPE computing unit are connected successively, and the output of frequency band SPE computing unit links to each other with fluctuation degree computing unit with minute frame SPE computing unit respectively;
Described spacing perception entropy judgement module comprises threshold value decision unit and fluctuation degree decision unit, divide the output of frame SPE computing unit to enter the threshold value decision unit, divide the output of frame SPE computing unit, fluctuation degree computing unit and threshold value decision unit to enter fluctuation degree decision unit;
The output of described spacing perception entropy judgement module and spatial parameter preliminary election delivery piece inserts the spatial parameter selection adjusting module.
Referring to Fig. 3, space parameter selection method based on spacing perception entropy judgement provided by the invention, the realization principle is carried out spatial parameter exactly and is chosen in advance, carrying out SPE and fluctuation degree thereof then calculates, carries out the SPE judgement according to SPE and fluctuation degree thereof, choose the adjustment of basic enterprising row space selection of parameter, the spatial parameter selection result who is optimized according to court verdict in advance at spatial parameter at last.As the technical scheme that can software realizes, except this basic procedure, space parameter selection method provided by the invention can also adopt multiple flow process to realize, for example carry out SPE and fluctuation degree thereof earlier and calculate, carry out the SPE judgement, carry out then carrying out the spatial parameter selection adjustment according to court verdict after spatial parameter is chosen in advance according to SPE and fluctuation degree thereof.An embodiment below is provided; with concrete parameter adjustment process as can invoked step; more meeting software design custom, is a kind of preferred version of space parameter selection method of the present invention when specifically implementing, also should be in protection domain of the presently claimed invention.This scheme is chosen spatial parameter from binaural cue ILD/ITD/IC, may further comprise the steps:
Step 1, employing Bark (BARK) band division methods are divided into 24 frequency bands with the full range band, obtain the scheme of choosing in advance of spatial parameter at the JND of full range band curve according to spatial parameter ILD/ITD/IC.Embodiment is:
(1) JND with single spatial parameter is an analytic target, compares the JND of nearby frequency bands, and the JND that numerical value is close is classified as one group.As shown in Figure 4, according to the JND curve of ILD, be boundaries of packets with frequency band 10, frequency band 15, frequency band 19, the full range band is divided into 4 groups.Adopting same quadrat method, at the JND curve of ITD, is boundaries of packets with frequency band 2, frequency band 18, frequency band 29, and the full range band is divided into 4 groups; At the JND curve of IC, be boundaries of packets with frequency band 2, frequency band 10, frequency band 16, frequency band 20, the full range band is divided into 5 groups, this step can be realized by the separate bands grouped element.
(2) get the boundaries of packets of ILD/ITD/IC in the separate bands grouping, promptly frequency band 2, frequency band 10, frequency band 15, frequency band 16, frequency band 18, frequency band 19 and frequency band 20 as the boundaries of packets of spatial parameter combine frequency band grouping, are finished the grouping of spatial parameter combine frequency band.In this step,, can on the basis of above-mentioned spatial parameter combine frequency band grouping, merge close boundaries of packets, reduce grouping number and every group of interior selection of parameter number if the low or available code check of coding/decoding system requirement complexity is less.This step can be realized by the combine frequency band grouped element.
(3) on this basis, compare the JND size of each spatial parameter ILD/ITD/IC between group, the less spatial parameter of JND is chosen in this group, otherwise will not choose.This step can realize by choosing pre-judgement unit.
Step 2, referring to Fig. 5, for the specific input signal in sampling back, be that a frame is done N point FFT conversion with N sampling point, can realize by the time-frequency conversion unit.Current frame signal is divided into 24 Bark bands as the i frame signal, and i is the frame label, can be realized by the frequency band division unit.Specifically can be divided into following four steps afterwards:
(1) carries out the calculating of spatial parameter ILD/ITD/IC according to the subband signal after the FFT conversion, can realize by the spatial parameter extraction unit.
ILD weighs by following formula:
ILD i ( b ) = 20 log 10 Σ k = k b k b + 1 - 1 | X l ( k ) | 2 Σ k = k b k b + 1 - 1 | X r ( k ) | 2 (formula 3)
Wherein b is the Bark tape label, k bAnd k B+1Be respectively the initial FFT spectral line label of Bark band b and b+1, X l(k) and X r(k) represent left and right acoustic channels k bar FFT spectral line respectively.ILD is the corresponding critical band spectrum energy of a left and right acoustic channels ratio, is unit with dB.
ITD weighs by following formula:
ILD i ( b ) = 1 k b + 1 - k b Σ k = k b k b + 1 - 1 ( arg X l ( k + 1 ) - arg X l ( k ) ) (formula 4)
- 1 k b + 1 - k b Σ k = k b k b + 1 - 1 ( arg X r ( k + 1 ) - arg X r ( k ) )
Wherein arg represents to get the argument of complex number main value.ITD is the left and right acoustic channels critical band group delay time difference.
IC weighs by following formula:
IC i ( b ) = | Re { Σ k = k b k b + 1 - 1 X l ( k ) X r * ( k ) } | Σ k = k b k b + 1 - 1 | X l ( k ) | 2 Σ k = k b k b + 1 - 1 | X r ( k ) | 2 (formula 5)
Wherein Re represents to get real.
(2) divide frequency band to estimate the value of the spacing perception entropy SPE of spatial parameter ILD/ITD/IC, can realize by frequency band SPE computing unit.
Particularly, the formula below the spe value of ILD in Bark band b pressed is weighed:
(formula 6)
Formula below the spe value of ITD in Bark band b pressed is weighed:
Figure G2009100610285D00121
(formula 7)
Formula below the spe value of IC in Bark band b pressed is weighed:
Figure G2009100610285D00122
(formula 8)
ILD wherein i(b), ITD i(b) and IC i(b) represent the parameter value of ILD/ITD/IC at i frame signal frequency band b respectively, Δ τ (b)/Δ λ (b)/Δ η (b) is respectively the JND of ILD/ITD/IC at frequency band b, determines by the audition test.The amplitude compression of α reflection sense of hearing perception generally gets 0.6.Log can any natural number be the end.Usually the computing machine acquiescence is the end with 2.
(3) SPE of space parameter I LD/ITD/IC in the i frame in the difference signal calculated input signal, but characterize with the aware space quantity of information of this each sampled point of frame inner average, and unit is the bit/ sample.Frame SPE computing unit was realized in available minute.
Particularly, the formula below the SPE value of ILD in the i frame pressed is weighed:
SPE ILD ( i ) = 1 N Σ b = 1 24 spe ILD ( b ) (formula 9)
Formula below the SPE value of ITD in the i frame pressed is weighed:
SPE ITD ( i ) = 1 N Σ b = 1 24 spe ITD ( b ) (formula 10)
Formula below the SPE value of IC in the i frame pressed is weighed:
SPE IC ( i ) = 1 N Σ b = 1 24 spe IC ( b ) (formula 11)
Wherein, N is the frame length of FFT conversion.
(4) entropy wave that calculates the SPE of space parameter I LD/ITD/IC in the i frame signal moves degree.Among the embodiment, this fluctuation degree not only SPE with present frame is relevant, and relevant with the SPE of all frames before, can be realized by fluctuation degree computing unit.
Particularly, adopt following computing formula:
σ α ( i ) = Σ b = 1 n ( spe i ( b ) - E ( spe k ) ) 2 , k = 1,2 , . . . , i (formula 12)
Wherein, α represents the spatial parameter ILD/ITD/IC that chooses, σ α(i) the accumulative total standard deviation of the SPE of expression i frame signal parameter alpha, spe i(b) be the spacing perception entropy of spatial parameter, E (spe at i frame frequency band b k) be the mathematical expectation of preceding i frame signal spacing perception entropy.
In conjunction with Fig. 6, step 3 and step 4 are described further, wherein the judgment part is realized respectively by threshold value decision unit and fluctuation degree decision unit, the concrete adjustment finished by the spatial parameter selection adjusting module:
Step 3, with the SPE of parameter I LD/ITD/IC in space in the step 2 ILD(i), SPE ITD(i) and SPE IC(i) compare with its corresponding threshold respectively.If SPE αGreater than 0 and at very big thresholding max αBelow, change step 4 over to and carry out the judgement of SPE fluctuation degree; If SPE αLess than 0, change step 5 over to and carry out the spatial parameter selection project setting; If SPE αAt very big thresholding max αMore than, change step 6 over to and carry out the spatial parameter selection project setting.Wherein, α represents the spatial parameter ILD/ITD/IC that chooses, when specifically implementing, and each spatial parameter ILD, ITD, IC very big thresholding max separately αCan be provided with respectively.
The fluctuation degree σ of space parameter I LD/ITD/IC in step 4, the comparison step two ILD(i), σ ITD(i) and σ IC(i), determine the maximum parameter of SPE fluctuation, be designated as α σ max, it is an a certain parameter among the ILD/ITD/IC; The SPE of space parameter I LD/ITD/IC in the comparison step two ILD(i), SPE ITD(i) and SPE IC(i), determine the parameter of SPE minimum, be designated as α SPEminIf α σ max≠ α SPEmin, then change step 6 over to, carry out the spatial parameter selection project setting; Otherwise, change step 7 over to.
Step 5, on i frame full range band, do not carry out choosing of this spatial parameter.
Step 6, on i frame full range band, all carry out choosing of this spatial parameter.
Step 7, the spatial parameter that keeps step 1 on i frame full range band are chosen scheme in advance.

Claims (5)

1. the space parameter selection method based on spacing perception entropy judgement is characterized in that, may further comprise the steps:
Step 1 is carried out spatial parameter and is chosen in advance, comprises following three steps,
Step 1.1 according to the fluctuation situation of each spatial parameter critical perception threshold value in each frequency band, is carried out the separate bands grouping of single spatial parameter;
Step 1.2 is carried out the combine frequency band grouping of all spatial parameters on the separate bands packet by packet basis of single spatial parameter, whether each spatial parameter is chosen respectively in the combine frequency band grouping adjudicated in advance;
Step 1.3, comprehensive pre-judgement result to all spatial parameters obtains spatial parameter and chooses scheme in advance;
Step 2 is carried out time-frequency conversion and frequency band division to every frame input signal, extracts each spatial parameter, calculates the spacing perception entropy and the moving degree of entropy wave thereof of each spatial parameter;
Step 3 compares the spacing perception entropy of each spatial parameter in the step 2 respectively with its corresponding threshold, carry out the spatial parameter selection project setting according to comparative result at every frame input signal, if spacing perception entropy is greater than 0 and at very big thresholding max αBelow, the moving degree of the spacing perception entropy of the different spaces parameter of more every frame input signal and entropy wave thereof if entropy wave moving degree maximum and spacing perception entropy are not minimum, all carries out choosing of this spatial parameter on the full range band; Otherwise the spatial parameter that keeps step 1 gained on the full range band is chosen scheme in advance;
If spacing perception entropy less than 0, does not carry out choosing of this spatial parameter on the full range band;
If spacing perception entropy is at very big thresholding max αMore than, all carry out choosing of this spatial parameter on the full range band.
2. space parameter selection method according to claim 1, it is characterized in that: in step 1.1, the separate bands of the carrying out single spatial parameter concrete mode of dividing into groups is, be analytic target at first with the critical perception threshold value of single spatial parameter in each frequency band, the critical perception threshold value that compares nearby frequency bands, the critical perception threshold value that numerical value is close is classified as one group, thereby finishes the separate bands grouping of single spatial parameter;
In step 1.2, the boundaries of packets of separate bands grouping of getting each spatial parameter is finished the combine frequency band grouping as the boundaries of packets of the combine frequency band grouping of all spatial parameters according to the boundaries of packets of combine frequency band grouping; Whether each spatial parameter is chosen the concrete mode of adjudicating in advance respectively in the combine frequency band grouping be, the critical perception threshold size of more single spatial parameter between group, spatial parameter to the less group of critical perception threshold value is chosen, otherwise will not choose.
3. space parameter selection method according to claim 1 is characterized in that: in step 2, when calculating moving the spending of entropy wave of present frame, with reference to the spacing perception entropy of present frame and the spacing perception entropy of frame before.
4. space parameter selection method according to claim 1 is characterized in that: the space perceptual entropy is greater than 0 and at very big thresholding max in step 3 αWhen following, at first the moving degree of the entropy wave of comparison step 2 each spatial parameter of gained is chosen the maximum spatial parameter of the moving degree of entropy wave; The spacing perception entropy that compares each spatial parameter then, the spacing perception entropy of the spatial parameter maximum as if the moving degree of entropy wave is not a minimum, then all carries out choosing of this spatial parameter on the full range band, otherwise the spatial parameter of reservation step 1 gained is chosen scheme in advance on the full range band.
5. a device that is used for claim 1 or 2 or 3 or 4 described space parameter selection methods is characterized in that: include the moving degree of spatial parameter preliminary election delivery piece, spacing perception entropy and entropy wave computing module, spacing perception entropy judgement module and spatial parameter selection adjusting module;
Described spatial parameter preliminary election delivery piece comprises separate bands grouped element, the combine frequency band grouped element that order links to each other and chooses pre-judgement unit;
The moving degree of described spacing perception entropy and entropy wave computing module comprises time-frequency conversion unit, frequency band division unit, spatial parameter extraction unit, frequency band SPE computing unit, divides frame SPE computing unit and fluctuation degree computing unit, time-frequency conversion unit, frequency band division unit, spatial parameter extraction unit and frequency band SPE computing unit are connected successively, and the output of frequency band SPE computing unit links to each other with fluctuation degree computing unit with minute frame SPE computing unit respectively;
Described spacing perception entropy judgement module comprises threshold value decision unit and fluctuation degree decision unit, divide the output of frame SPE computing unit to enter the threshold value decision unit, divide the output of frame SPE computing unit, fluctuation degree computing unit and threshold value decision unit to enter fluctuation degree decision unit;
The output of described spacing perception entropy judgement module and spatial parameter preliminary election delivery piece inserts the spatial parameter selection adjusting module;
Wherein, SPE representation space perceptual entropy.
CN2009100610285A 2009-03-09 2009-03-09 Spacing parameter choosing method and apparatus based on spacing perception entropy judgement Expired - Fee Related CN101499280B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2009100610285A CN101499280B (en) 2009-03-09 2009-03-09 Spacing parameter choosing method and apparatus based on spacing perception entropy judgement

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2009100610285A CN101499280B (en) 2009-03-09 2009-03-09 Spacing parameter choosing method and apparatus based on spacing perception entropy judgement

Publications (2)

Publication Number Publication Date
CN101499280A CN101499280A (en) 2009-08-05
CN101499280B true CN101499280B (en) 2011-11-02

Family

ID=40946326

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2009100610285A Expired - Fee Related CN101499280B (en) 2009-03-09 2009-03-09 Spacing parameter choosing method and apparatus based on spacing perception entropy judgement

Country Status (1)

Country Link
CN (1) CN101499280B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102760442B (en) * 2012-07-24 2014-09-03 武汉大学 3D video azimuth parametric quantification method
CN102867518B (en) * 2012-09-10 2014-07-02 武汉大学 Encoding-decoding performance evaluating method for horizontal orientation parameters in 3D (three-dimensional) audio
CN107665711A (en) * 2016-07-28 2018-02-06 展讯通信(上海)有限公司 Voice activity detection method and device
CN109979476B (en) * 2017-12-28 2021-05-14 电信科学技术研究院 Method and device for removing reverberation of voice

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101162904A (en) * 2007-11-06 2008-04-16 武汉大学 Space parameter stereo coding/decoding method and device thereof
CN101282479A (en) * 2008-05-06 2008-10-08 武汉大学 Method for encoding and decoding airspace with adjustable resolution based on interesting area

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101162904A (en) * 2007-11-06 2008-04-16 武汉大学 Space parameter stereo coding/decoding method and device thereof
CN101282479A (en) * 2008-05-06 2008-10-08 武汉大学 Method for encoding and decoding airspace with adjustable resolution based on interesting area

Also Published As

Publication number Publication date
CN101499280A (en) 2009-08-05

Similar Documents

Publication Publication Date Title
US8082157B2 (en) Apparatus for encoding and decoding audio signal and method thereof
US20080201152A1 (en) Apparatus for Encoding and Decoding Audio Signal and Method Thereof
EP3605847B1 (en) Multichannel signal encoding method and apparatus
EP3493203B1 (en) Method for encoding multi-channel signal and encoder
TWI404429B (en) Method and apparatus for encoding/decoding multi-channel audio signal
US20080249769A1 (en) Method and Apparatus for Determining Audio Spatial Quality
CN101149925B (en) Space parameter selection method for parameter stereo coding
CN101499280B (en) Spacing parameter choosing method and apparatus based on spacing perception entropy judgement
US20240135938A1 (en) Multi-Channel Signal Encoding Method, Multi-Channel Signal Decoding Method, Encoder, and Decoder
CN101297352A (en) Apparatus for encoding and decoding audio signal and method thereof
CN106231528A (en) Personalized head related transfer function based on stagewise multiple linear regression generates system and method
CN101427307A (en) Method and apparatus for encoding/decoding multi-channel audio signal
US20110040566A1 (en) Method and apparatus for encoding and decoding residual signal
CN101521013B (en) Spatial audio parameter bidirectional interframe predictive coding and decoding devices
CN105575401A (en) AACHuffman domain steganalysis method based on C-MAC characteristics
CN101504835B (en) Measurement method for spacial sensed information content in acoustic field and application thereof
Chen et al. Interpolation method of head-related transfer functions based on common-pole/zero modeling
Cao et al. Spatial parameter choosing method based on spatial perception entropy judgment
Arbel et al. On HRTF Notch Frequency Prediction using Anthropometric Features and Neural Networks
Wang et al. The perceptual and statistics characteristic of spatial cues and its application
JP2022505964A (en) Directional volume map based audio processing
CN104205211A (en) Multi-channel audio encoder and method for encoding a multi-channel audio signal

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20111102

Termination date: 20150309

EXPY Termination of patent right or utility model