WO2009121263A1

WO2009121263A1 - Filtering method and filter

Info

Publication number: WO2009121263A1
Application number: PCT/CN2009/070813
Authority: WO
Inventors: · 鲁尔门海因里希·W; 皮特·瓦里; 霍克·克鲁格; 本特·凯瑟; 张德明; 张琦
Original assignee: 华为技术有限公司
Priority date: 2008-03-29
Filing date: 2009-03-16
Publication date: 2009-10-08
Also published as: CN101546992A; CN101546992B

Abstract

A filtering method and filter..The method includes: receiving a first filtered signal processed by low-pass filtering and a second filtered signal processed by high-pass filtering,all-pass filtering a third filtered signal for the phase equalizing, all-pass filtering a fourth filtered signal for the phase equalizing, synthesizing the third and fourth filtered signal which are all-pass filtered for the phase equalizing, and accepting the output signal.The filter includes: a synthesis filter group,which comprises an accepting module,a first all-pass filtering module, a second all-pass filtering module and a synthesis process module. Thus the amplitude distortion and the phase distortion can be eliminated, and the computational complexity and the arithmetic time delay can be also reduced..

Description

The present invention claims the priority of the Chinese patent application filed on March 29, 2008, the Chinese Patent Application No. 200810087037.7, entitled "A Filter Method and Filter", the entire contents of which is incorporated herein by reference. This is incorporated herein by reference. Technical field

The embodiments of the present invention relate to the field of signal processing, and in particular, to a filtering method and a filter. Background technique

Quadrature Mirror Filter (QMF) is widely used in signal processing, especially in the fields of speech coding, digital video, communication systems and short-term spectrum analysis. A variety of QMF structures and filtering methods are provided in the prior art, for example: Infinity Impulse Response (IIR)/Finity Impulse Response (FIR) Hybrid Filter is used to achieve approximation Fully reconstructed QMF (NPR-QMF) for approaching PR-QMF. In this scheme, the IIR filter is used to implement band division and combination, and the FIR filter is used to compensate the phase distortion caused by the IIR filter. The FIR filter for phase compensation is designed as:

^ ( 1 ) where "' and the filter parameters, the phase-compensated analysis filter bank transfer function is:

Γ« (ζ) = Η _αι (ζ) . « (ζ) = - (ζ - a _; )∑z"- ^d '(α,.)" = ζ^'

z ~ ^a i ( 2 ) This scheme solves the problem of phase distortion introduced by IIR, but there is still some amplitude distortion, and phase compensation using FIR filter will introduce higher computational complexity. Summary of the invention

Embodiments of the present invention provide a filtering method and a filter, which are used to completely eliminate amplitude distortion and aliasing distortion, approximate a linear phase of an ideal state, and thereby eliminate phase distortion; and, reduce computational complexity and reduce algorithm time Delay.

The embodiment of the invention provides a filtering method, including:

Receiving a first filtered signal subjected to low pass filtering processing and a second filtered signal subjected to high pass filtering processing;

Obtaining a third filtered signal according to the first filtered signal and the second filtered signal, performing phase-equalized all-pass filtering processing on the third filtered signal;

Obtaining a fourth filtered signal according to the first filtered signal and the second filtered signal, and performing a phase-equalized all-pass filtering process on the fourth filtered signal;

The third filtered signal subjected to phase-equalized all-pass filtering processing and the fourth filtered signal are combined to obtain an output signal.

Embodiments of the present invention provide a filter, including a synthesis filter bank, where the synthesis filter bank includes:

a receiving module, configured to receive a first filtered signal that has undergone low-pass filtering processing and a second filtered signal that has undergone high-pass filtering processing;

a first all-pass filtering module, configured to acquire a third filtered signal according to the first filtered signal and the second filtered signal, and perform phase-equalized all-pass filtering processing on the third filtered signal;

a second all-pass filtering module, configured to acquire a fourth filtered signal according to the first filtered signal and the second filtered signal, and perform phase-equalized all-pass filtering processing on the fourth filtered signal;

And a synthesis processing module, configured to synthesize the third filtered signal and the fourth filtered signal subjected to phase-equalized all-pass filtering to obtain an output signal.

In the filtering method and the filter provided by the embodiment of the invention, the first filtered signal is subjected to low-pass filtering processing, and the second filtered signal is subjected to high-pass filtering processing, which can eliminate aliasing distortion and amplitude distortion, and at the same time, the third filtered signal and the fourth filtering The signal is processed by phase-equalized all-pass filtering, approaching the ideal The linear phase of the state, thereby eliminating phase distortion; and, in the embodiment of the present invention, the computational complexity is reduced, and the delay of the algorithm is reduced. DRAWINGS

1 is a flowchart of a filtering method according to an embodiment of the present invention;

2 is a schematic diagram of a basic structure of a QMF in an application scenario of a filtering method according to an embodiment of the present invention; FIG. 3 is an application scenario of a filtering method according to an embodiment of the present invention. The ideal amplitude-frequency response curve of (and ι ( ^ζ );

4 is a schematic diagram of a QMF multi-phase structure that satisfies the condition of no aliasing distortion in an application scenario in a filtering method according to an embodiment of the present invention;

FIG. 5 is a schematic structural diagram of a system of a QMF in a filtering method according to Embodiment 2 of the present invention; FIG.

FIG. 6 is a schematic diagram of a filter according to an embodiment of the present invention. detailed description

The technical solutions of the embodiments of the present invention are further described in detail below with reference to the accompanying drawings and embodiments.

As shown in FIG. 1 , a flowchart of a filtering method according to an embodiment of the present invention includes the following steps: Step 101: Receive a first filtered signal that has undergone low-pass filtering processing and a second filtered signal that has undergone high-pass filtering processing;

Step 102: Acquire a third filtered signal according to the first filtered signal and the second filtered signal, and perform a phase-equalized all-pass filtering process on the third filtered signal;

Step 103: Acquire a fourth filtered signal according to the first filtered signal and the second filtered signal, and perform a phase-equalized all-pass filtering process on the fourth filtered signal;

Step 104: Synthesize the third filtered signal and the fourth filtered signal subjected to phase-equalized all-pass filtering to obtain an output signal.

Step 104 may be specifically: performing third filtering of phase-equalized all-pass filtering processing The signal and the fourth filtered signal are subjected to interpolation processing; the third filtered signal subjected to the interpolation processing is subjected to delay processing, and the sum of the third filtered signal subjected to the delay processing and the fourth filtered signal subjected to the interpolation processing is obtained as an output signal.

The foregoing step 101 may further include: Step 100: Acquire the first filtered signal and the second filtered signal according to the input signal. The step may specifically include:

Step 1001: Perform an extraction process on the input signal and the input signal subjected to the delay processing respectively; Step 1002: Perform an all-pass filtering process on the first input signal by using the input signal subjected to the decimation process as the first input signal;

Step 1003: The input signal subjected to the delay processing and the decimation processing is used as a second input signal, and the second input signal is subjected to all-pass filtering processing;

Step 1004, calculating a sum of the first input signal and the second input signal subjected to the all-pass filtering process, and acquiring the first filtered signal according to the sum of the first input signal and the second input signal, and transmitting;

Step 1005: Calculate a difference between the first input signal and the second input signal processed by the all-pass filter, and obtain a second filtered signal according to a difference between the first input signal and the second input signal and send the second filtered signal.

Further, in step 1004, acquiring the first filtered signal according to the sum of the first input signal and the second input signal is specifically: encoding the sum of the first input signal and the second input signal to obtain a first filtered signal; In 1005, acquiring the second filtered signal according to the difference between the first input signal and the second input signal is: performing a coding process on the difference between the first input signal and the second input signal to obtain a second filtered signal.

Then, in step 102, acquiring the third filtered signal according to the first filtered signal and the second filtered signal is: decoding the first filtered signal and the second filtered signal, and calculating the first filtered signal after decoding processing The sum of the two filtered signals is used as the third filtered signal. In step 103, acquiring the third filtered signal according to the first filtered signal and the second filtered signal is specifically: decoding the first filtered signal and the second filtered signal, and calculating The difference between the processed first filtered signal and the second filtered signal is decoded as a fourth filtered signal.

In this embodiment, the first filtered signal is subjected to low-pass filtering processing, and the second filtered signal is subjected to high-pass filtering. Wave processing can eliminate aliasing distortion and amplitude distortion. At the same time, the third filtered signal and the fourth filtered signal are subjected to phase-equalized all-pass filtering to approximate the linear phase of the ideal state, thereby eliminating phase distortion; and, the implementation The example filtering method reduces the computational complexity and reduces the delay of the algorithm.

The filtering method of the embodiment of the present invention takes the filtering method of the QMF as an example to further introduce the technical solution of the embodiment of the present invention.

An application scenario of the filtering method in the embodiment of the present invention is first introduced. As shown in FIG. 2, it is a schematic diagram of a QMF basic structure in an application scenario of a filtering method according to an embodiment of the present invention. The QMF divides the input signal x(n) into M (M= 2) Sub-band signals, the bandwidth of the M sub-band signals is 1/M of the original. Add a M (M = 2) times decimator 11 after the analysis filter bank (including ( ^z ) and (z)) at the transmitting end; synthesis filter bank at the receiving end (including ^. (and ^^) ) An interpolator 12 of M (M = 2) times is added before.

The transfer function of the above QMF system can be expressed as:

Y(z) = X(z)T _lin (z) + X(-z)E _alias (z)

Its towel:

[H ₀ (z)G ₀ (z) + H , (2)0, (2)]

(4)

E _aKas (z) = -[Ho (-z)G ₀ (z) + H _x (-2)0, (z)]

(5) Equation (4) represents the linear transfer function of the QMF system, and equation (5) represents the alias transfer function of the QMF system.

If the above synthesis filter bank is designed as:

G ₀ (z) = C{z)H, (-z), G, (z) = -C(z)H ₀ (-z) _{( 6 ) The} reconstructed output signal has no aliasing distortion, ie:

Υ(ε ^]ω ) = X{e ^j0} )T _Hn {e ^j0} ( _η ) If the above analysis filter bank is designed to meet

, then the reconstructed output signal has no amplitude Degree distortion, ie:

If the above analysis filter bank is designed

Can satisfy +, then the reconstructed output signal has no phase distortion, namely:

Arg{7(^ ^ffl )} = arg{ (e^)} + αω + β ( ₉ ) As shown in FIG. 3 , it is an application scenario in the filtering method of the embodiment of the present invention. The ideal amplitude-frequency response curve of (and ^(). QMF that completely eliminates aliasing distortion, amplitude distortion, and phase distortion can be called fully reconstructed QMF (PR-QMF).

The QMF can also be implemented by using a multi-phase structure. As shown in FIG. 4, it is a schematic diagram of a QMF multi-phase structure that satisfies the condition of no aliasing distortion in an application scenario in the filtering method of the embodiment of the present invention. In this structure,

H ₀ ( _Z ) = E. (z ² ) + z- 0 ² ), H ₁ (z)^E ₀ (z ² )-z- ¹ E ₁ (z ² ) _? T(z)^2z- ^l E ₀ (z ² )E _l The nature of (z ² ) _{( l0 )} determines whether there is amplitude distortion and phase distortion.

As shown in FIG. 5, it is a schematic diagram of a system structure of a QMF in a filtering method according to a second embodiment of the present invention, which includes an analysis filter group 2 and a synthesis filter bank 3.

In the analysis filter bank 2, the input signal first passes through the multiplier 21, and then one input signal is sent to the decimator 22 for decimation processing, and the other input signal is subjected to delay processing and then sent to the decimator 23 for decimation processing; The input signal is used as the first input signal, and the input signal subjected to the delay processing and the decimation processing is used as the second input signal, and the first input signal and the second input signal are respectively performed in the all-pass filter 24 and the all-pass filter 25, respectively. All-pass filtering processing; calculating the sum of the first input signal and the second input signal subjected to the all-pass filtering process as the first filtered signal in the adder 26; calculating the first input subjected to the all-pass filtering process in the adder 27 The difference between the signal and the second input signal is used as the second filtered signal.

In the synthesis filter bank 3, first, the sum of the received first filtered signal and the second filtered signal is calculated in the adder 31 as a third filtered signal, and the first filtered signal is calculated at the adder 32. a difference between the second filtered signal and the second filtered signal; the third filtered signal and the fourth filtered signal are all-pass filtered in the all-pass filter 33 and the all-pass filter 34, respectively; The processed third filtered signal and the fourth filtered signal are respectively subjected to interpolation processing in the interpolator 35 and the interpolator 36; after the third filtered signal subjected to the interpolation processing is subjected to delay processing, the third filtered signal is calculated in the adder 37 and The sum of the fourth filtered signals is used as an output signal; the output signal is output by the multiplier 38.

Wherein, the all-pass filtering formulas of the two all-pass filters in the analysis filter bank 2 are respectively represented by a sum, and then the transfer function ^ of the entire analysis filter bank 2 is analyzed. The expressions of and ( ^z ) are:

H ₀ (z)=^[H _a0 (z ² ) + z ¹ H _al (z ² )]

(The filtering formulas of the two all-pass filters in the 12 J synthesis filter bank 3 are respectively used. (And A( ^z ), then the transfer function ^{G of} the entire synthesis filter bank 3. ( ^z ^ ^G i( ^z ) The expressions are:

G ₀ (z) = ^[z ^l B ₀ (z ² ) + B _l (z ² )]

( 13 ) = [ — _{( 14 )} Therefore, the linear transfer function and the alias transfer function of the entire QMF system are:

T _lin (^) - ^Z a _a , (^ ² )B ₀ (z ² ) + H _al (z ² )B, (z ² )] ( _{i5 )}

E _alias (^) = ^-[ _a0 (z ² )B ₀ (z ² )-H _al {z )B _x (z ² )] ( ^ ) This embodiment will analyze two all-passes in filter bank 2 The filter's all-pass filtering formula is designed as follows:

i -1<« _; <1;« _; ei?; e{0,l} (18) where is the analysis filter parameter. The all-pass filtering process of the first input signal may be specifically: multiplying the first input signal by the equation (17), at this time, i = 0; performing the all-pass filtering process on the second input signal may be: The second input signal is multiplied by equation (17), at which point i = 1.

The above equation shows that ⁷ ^( ^z ) is a first-order all-pass filter, derived from this, ^. (It is an IIR low-pass filter, ( ^z ) is an IIR high-pass filter, that is, the first filtered signal output by the analysis filter bank 2 is subjected to low-pass filtering processing, and the second filtered signal is subjected to high-pass filtering processing.

The above analysis filter bank 2 introduces phase distortion, so when designing the synthesis filter 3, it is considered to introduce a stable all-pass filter for phase equalization to reduce the phase distortion problem. The present embodiment designs the synthesis filter bank 3 based on an all-pass filter capable of achieving phase equalization.

An all-pass filter that achieves phase equalization can be expressed as:

The filter shown in equation (19) may also be referred to as a phase equalizer, where the order of the phase equalizer is represented, the upper limit of which is limited by the specific implementation environment. According to the phase equalizer, the filtering formula of the two all-pass filters in the synthesis filter bank 3 is designed as follows:

Among them, ^. And the delay parameter, d _{i =} 2 ^N - ; Equation (20) is the first filtering formula, and Equation (21) is the second filtering formula. The all-pass filtering process for performing phase equalization on the third filtered signal is specifically: multiplying the third filtered signal by the equation (20); and performing all-pass filtering processing for phase equalizing the fourth filtered signal is: The signal is multiplied by equation (21).

In order not to lose generality, suppose ^^^ and ^ ^. It can be seen from equations (20) and (21) that the parameters of the phase-equalized all-pass filter are determined by the parameters of the analysis filter bank 2. Substituting the formula (20) and the formula (21) into the formula (15), and according to the formula (19), can be obtained:

T (z) = - ⁰ +— ^z (22 )

2 z ^2d ° - « ₀ ^d ° 卞 2 z ^2di - ^Known by equation (22), with the delay parameter ^. The increase of sum, the above linear transfer function tends to be delayed

+ l delay unit.

Substituting the formula (20) and the formula (21) into the formula (16), and according to the formula (19), can be obtained:

-2(d,-d ₀ )-ld ₀ 2d ₀ -1 d 2dl

_F ( ₇ \ ₌ ^L i- ₀ zz ia _l z ) αΐύΛ 2 _z ^2d . _« ₀ ^d . 2 z ^2di -af From equation (23), the aliasing distortion can be arbitrarily small when the sum is large enough.

Observe equation (19). When calculating, you need to perform ²¹ § 2 real multiplications, ²¹ § 2 real additions, and ¹ delay. Observing the background techniques (1) and (2), you need to perform submultiplication and sub-addition. , - ¹ time delay, therefore, compared with the prior art, the computational complexity is reduced, the delay of the algorithm is reduced, and the computational complexity and algorithm delay of the QMF system are determined by the order of the phase equalizer, The smaller the order, the lower the computational complexity and the smaller the algorithm delay; since the filter bank 2 is analyzed in this embodiment. (IIR low-pass filter, ^(IRR high-pass filter, so aliasing distortion and amplitude distortion can be eliminated, a stable all-pass filter is introduced in synthesis filter bank 3 for phase equalization, approximating linear phase, thus eliminating The phase distortion is determined, and the degree of phase distortion of the QMF system is determined by the order of the phase equalizer. The larger the order, the closer to the linear phase; therefore, the present embodiment can flexibly control the entire QMF by setting the order of the phase equalizer. The degree of phase distortion, computational complexity, and algorithm delay of the system.

The third filtering method of the embodiment of the present invention can design the all-pass filtering formula of two all-pass filters in the analysis filter bank as follows: a, {z = - ¹ ^- z - « _; (24)

|α,.|<|ζ|;-1<«_;<1;«_;ei?; e{0,l} (25) This is the same as in the second embodiment. The all-pass filtering process of the first input signal may be specifically as follows: The first input signal is multiplied by the equation (24), and i = 0; the second input signal is subjected to the all-pass filtering process, which may be specifically: multiplying the second input signal by the equation (25), , i= 1

The difference between this embodiment and the second embodiment is that the filtering formulas of the two all-pass filters in the synthesis filter bank are designed as follows:

B _l (z) = 2P^(z)T^(z) (27) Equation (26) is the third filtering formula, and equation (27) is the fourth filtering formula. The all-pass filtering process for performing phase equalization on the third filtered signal is specifically: multiplying the third filtered signal with the equation (26); and performing all-pass filtering processing for phase equalizing the fourth filtered signal is: The signal is subjected to a product operation of equation (27).

Where, calculate ⁷ and ⁷ ^ according to the following formula

Since -1<« _; <1 , Equation ( ₂₈ ) also represents a stable all-pass filter in which the parameters of the analysis filter bank are considered, and when increased, the filter characteristics tend to a single delay unit. ^d '. Substituting equations (26) and (27) into equation (15), and according to equations (19) and (28), can be obtained:

1 _ . 1- ,

γ ( \ C 0 )

_ _{a 0} according to equations (26), (27), (16), (19) and (28):

E _a!ias (z) = Q (30) Therefore the transfer function of the entire QMF system is a ² ( +^ ^{+ 1st} order all-pass filter, so

The QMF system has no amplitude distortion and aliasing distortion, and as the sum of 4 increases, the entire system tends to be linear, and phase distortion can be eliminated.

In this embodiment, the phase equalizer is used to perform the level with an all-pass filter that considers the parameters of the analysis filter bank. Compared with the prior art, the two all-pass filters of the synthesis filter are reduced, and the computational complexity is reduced, and the delay of the algorithm is reduced; since the filter bank is analyzed in this embodiment. (IIR low-pass filter, ^(IRR high-pass filter, so aliasing distortion and amplitude distortion can be eliminated, a stable all-pass filter is introduced into the synthesis filter bank for phase equalization, approximating the linear phase, thus eliminating Phase distortion.

The fourth filtering method of the embodiment of the present invention can design the transfer function of the analysis filter bank as follows:

H. (z) = [Γ (ζ ² ) + ζ- ^ι Η _αι (ζ ² (ζ ² )] (31) ) — ζ- ^ι Η _αι (ζ ² ) ) (ζ ² )] (32)

Where equation (31) is the first analytical filtering formula and equation (32) is the second analytical filtering formula. Obtaining the first filtered signal according to the input signal is specifically: multiplying the input signal by the equation (31) to obtain the first filtered signal; and acquiring the second filtered signal according to the input signal, specifically: multiplying the input signal by the equation (32) Operation, obtaining a second filtered signal.

The transfer function of the synthesis filter bank is designed as follows:

Where equation (33) is the first synthesis filter formula and equation (34) is the second synthesis filter formula. The third filtered signal is subjected to a product operation of equation (33) to obtain an output signal; and the fourth filtered signal is subjected to a product operation of equation (34) to obtain an output signal.

The above formulas (31) to (34) can also be expressed as:

H;(z) = H„ ) (35) G _i '(z) = 2G _i (z)P^(z ² ) (36) where e{0,l}. Since ^)(z ² ) is used For phase-balanced all-pass filters, so H:( ^z ) and

^H ' ^z and (the same as the amplitude-frequency response, and the whole system tends to be more linear. This embodiment has no amplitude distortion and aliasing distortion, and the phase distortion is also smaller; and, compared with the prior art Compared, the computational complexity is reduced and the algorithm delay is reduced.

In the filtering method of the embodiment of the present invention, if ⁷ ^( ^Z ) is a high-order all-pass filter, it can be represented as a product of a plurality of first-order all-pass filter transfer functions, and the subsequent processing methods are the same.

As shown in FIG. 6 , which is a schematic diagram of a filter according to an embodiment of the present invention, the filter includes a synthesis filter bank 4, wherein the synthesis filter bank 4 includes: a receiving module 41, configured to receive a first filter that is subjected to low-pass filtering processing. And a second filtered signal processed by the high-pass filter; the first all-pass filtering module 42 is configured to obtain a third filtered signal according to the first filtered signal and the second filtered signal, and perform phase equalization on the third filtered signal An all-pass filtering process; the second all-pass filtering module 43 is configured to obtain a fourth filtered signal according to the first filtered signal and the second filtered signal, and perform phase-equalized all-pass filtering processing on the fourth filtered signal; The synthesis processing module 44 is configured to synthesize the third filtered signal and the fourth filtered signal that have undergone phase equalization all-pass filtering processing to obtain an output signal.

The synthesis processing module may include: an interpolation module, configured to perform interpolation processing on the third filtered signal subjected to phase-equalized all-pass filtering processing and the fourth filtered signal; and an output module, configured to perform interpolation processing The third filtered signal performs delay processing, and obtains a sum of the third filtered signal subjected to the delay processing and the fourth filtered signal subjected to the interpolation processing as an output signal.

The embodiment may further include an analysis filter bank 5 configured to acquire the first filtered signal and the second filtered signal according to an input signal. Further, the analysis filter bank 5 may include: a decimation module 51 for performing decimation processing on the input signal and the delay-processed input signal, respectively; and a third all-pass filtering module 52 for using the demodulated input signal as a first input signal, performing an all-pass filtering process on the first input signal; a fourth all-pass filtering module 53 configured to use the input signal subjected to the delay processing and the decimation processing as a second input signal, and the second input The signal is subjected to an all-pass filtering process; the first adder 54 is configured to calculate a sum of the first input signal and the second input signal subjected to the all-pass filtering process, and obtain the sum of the first input signal and the second input signal according to the sum of the first input signal and the second input signal The first filtered signal is sent and sent; the second adder 55 is configured to calculate the first processed by the all-pass filter a difference between an input signal and a second input signal, and acquiring the second filtered signal according to a difference between the first input signal and the second input signal and transmitting.

In this embodiment, the first filtered signal is subjected to low-pass filtering processing, and the second filtered signal is subjected to high-pass filtering processing to eliminate aliasing distortion and amplitude distortion, and at the same time, the third filtered signal and the fourth filtered signal are all phase-balanced all-pass The filtering process approximates the linear phase of the ideal state, thereby eliminating phase distortion; and, in this embodiment, the computational complexity of the filter is reduced, and the delay of the algorithm is reduced.

A person skilled in the art can understand that all or part of the steps of implementing the above method embodiments may be completed by using hardware related to program instructions, and the foregoing program may be stored in a computer readable storage medium, and the program is executed when executed. The steps of the foregoing method embodiments are included, and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk. It is not limited thereto; although the embodiments of the present invention have been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or some of the technologies. The features are equivalent to the equivalents of the technical solutions of the embodiments of the embodiments of the present invention.

Claims

Rights request

A filtering method, comprising:

The filtering method according to claim 1, wherein the receiving the first filtered signal subjected to the low-pass filtering process and the second filtering signal subjected to the high-pass filtering process further comprises: acquiring the a first filtered signal and the second filtered signal.

The filtering method according to claim 2, wherein the acquiring the first filtered signal and the second filtered signal according to the input signal comprises:

Extracting the input signal and the delayed input signal separately;

The decimated input signal is used as a first input signal, and the first input signal is subjected to all-pass filtering processing;

The input signal subjected to the delay processing and the decimation processing is used as a second input signal, and the second input signal is subjected to all-pass filtering processing;

Calculating a sum of the first input signal and the second input signal subjected to the all-pass filtering process, and acquiring the first filtered signal according to the sum of the first input signal and the second input signal and transmitting;

Calculating a difference between the first input signal and the second input signal subjected to the all-pass filtering process, and acquiring the second filtered signal according to a difference between the first input signal and the second input signal and transmitting.

The filtering method according to claim 3, wherein the performing the all-pass filtering process on the first input signal/the second input signal is specifically: the first input signal/second input signal The number is multiplied by the all-pass filter formula, and the all-pass filter formula is represented by »:

Wherein the filtering parameter is analyzed, l ^a 'l<l ^z l , when the first input signal is multiplied by the all-pass filtering formula, i = 0; when the second input signal is multiplied by the all-pass filtering formula When calculating, i= 1.

The filtering method according to claim 4, wherein the all-pass filtering process of performing phase equalization on the third filtered signal is specifically: performing a product operation on the third filtered signal and the first filtering formula, The first filtering formula is represented by A( ^z ):

B ₀ (z) = 2P^(z)-z- ^(d '- ^do)

Where, and 4 are delay parameters;

The all-pass filtering process for performing phase equalization on the fourth filtered signal is specifically: performing a product operation on the fourth filtered signal and a second filtering formula, where the second filtering formula is represented by A(z):

Calculate ^ and ^) according to the following formula: z + a,

Where ^ represents the order of the phase equalization filter.

The filtering method according to claim 4, wherein the all-pass filtering process of performing phase equalization on the third filtered signal is specifically: performing a product operation on the third filtered signal and the third filtering formula; The third filter formula uses (to indicate:

The all-pass filtering process for performing phase equalization on the fourth filtered signal is specifically: multiplying the fourth filtered signal and the fourth filtering formula, wherein the fourth filtering formula is represented by A(z):

B _l (z) = 2P^(z)T^(z) Calculate ^ and ^) according to the following formula:

l + («,.z) where, represents the order of the phase equalization filter;

Calculate ^^) and ⁷ ^ according to the following formula

Where di = 2 ^N '.

The filtering method according to claim 3, wherein the performing all-pass filtering processing on the first input signal is specifically: performing all-pass filtering processing for phase equalization of the first input signal; The all-pass filtering process of the input signal is specifically: performing all-pass filtering processing of phase equalization of the second input signal.

The filtering method according to claim 2, wherein the acquiring the first filtered signal and the second filtered signal according to the input signal is specifically: performing the input signal and a first analysis filtering formula a product operation, obtaining a first filtered signal; performing a product operation on the input signal and a second analysis filter formula to obtain a second filtered signal;

The first analysis filter formula is used to represent:

Hey. (ζ) = (ζ ² ) + ζ- ^ι Η _αι (ζ ² )Ρ^ (ζ ² )]

The second analysis filter formula is represented by H; ( _z ):

(z) = [Γ (ζ ² ) - ζ- ^ι Η _αι (ζ ² ) ) (ζ ² )] Calculate ^ (ζ ² ) according to the following formula:

Calculated according to the following formula

Where ^ represents the order of the phase equalization filter; ^^ ² ) is calculated according to the following formula:

Where = 2

The filtering method according to claim 8, wherein the all-pass filtering process of phase-equalizing the third filtered signal and the fourth filtered signal, and the all-pass filtering process of phase-equalizing The third filtered signal and the fourth filtered signal are combined to obtain an output signal, which is specifically:

And multiplying the third filtered signal and the first synthesis filter formula to obtain an output signal; the first synthesis filter formula uses G. (z) to indicate:

G ₀ (z) = z- ^l T^ (z ² ) + H _ao (z ² )P^ (z ² )

Or performing a product operation on the fourth filtered signal and the second synthesis filter formula to obtain an output signal; the second synthesis filter formula is represented by 0):

(ζ ² ).

The filtering method according to claim 1, wherein the acquiring the third filtered signal according to the first filtered signal and the second filtered signal is: calculating a sum of the first filtered signal and the second filtered signal, And acquiring the fourth filtered signal according to the first filtered signal and the second filtered signal, specifically: calculating a difference between the first filtered signal and the second filtered signal to obtain the fourth filtered signal.

A filter comprising a synthesis filter bank, wherein the synthesis filter bank comprises:

a receiving module, configured to receive a first filtered signal that has undergone low-pass filtering processing and a second filtered signal that has undergone high-pass filtering processing; a first all-pass filtering module, configured to acquire a third filtered signal according to the first filtered signal and the second filtered signal, and perform phase-equalized all-pass filtering processing on the third filtered signal;

The filter according to claim 11, wherein the synthesis processing module comprises:

An interpolation module, configured to perform interpolation processing on the third filtered signal subjected to phase-equalized all-pass filtering processing and the fourth filtered signal;

And an output module, configured to perform delay processing on the third filtered signal subjected to the interpolation processing, and obtain a sum of the third filtered signal subjected to the delay processing and the fourth filtered signal subjected to the interpolation processing as an output signal.

The filter according to claim 11, further comprising: an analysis filter bank, configured to acquire the first filtered signal and the second filtered signal according to an input signal.

The filter according to claim 13, wherein the analysis filter bank comprises:

a decimation module, configured to perform an extraction process on the input signal and the delayed input signal separately;

a third all-pass filter module, configured to perform the all-pass filtering process on the first input signal by using the input signal subjected to the decimation process as a first input signal;

a fourth all-pass filtering module, configured to perform an all-pass filtering process on the second input signal by using an input signal subjected to delay processing and decimation processing as a second input signal;

a first adder, configured to calculate a sum of the first input signal and the second input signal subjected to the all-pass filtering process, and obtain the first filtered signal according to a sum of the first input signal and the second input signal and send the first filtered signal ; a second adder, configured to calculate a difference between the first input signal and the second input signal processed by the all-pass filter, and obtain the second filtered signal according to a difference between the first input signal and the second input signal and send the second filtered signal .