P. 1
09-SubbandCoding

# 09-SubbandCoding

|Views: 17|Likes:

See more
See less

08/11/2011

pdf

text

original

# Subband and wavelet coding

                

Vector convolution, convolutional transforms Filter banks vs. vector space interpretation Orthogonal and biorthogonal subband transforms DCT as a ﬁlter bank Lapped Orthogonal Transform (LOT) Discrete Wavelet Transform (DWT) Quadrature mirror ﬁlters and conjugate quadrature ﬁlters Lifting Embedded zero-tree coding of wavelet coefﬁcients

Bernd Girod: EE398A Image and Video Compression

Subband and Wavelet Coding no. 1

Subband coding: motivation






Coding with block-wise transform introduces visible blocking artifacts, as bit-rate decreases. Can we, somehow, overlap adjacent blocks,   thereby smoothing block boundaries,   but without increasing the number of transform coefﬁcients? Solution: subband transform.

Bernd Girod: EE398A Image and Video Compression

Subband and Wavelet Coding no. 2

Vector convolution

+

+

Forward transform Inverse transform
Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. 3

4 .Perfect reconstruction condition   Original domain   z-transform: “polyphase matrices”   Perfect reconstruction condition in the z-domain   Example. m=2 Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

5 .Filter bank interpretation of convolutional transform Analysis ﬁlterbank subband signals Synthesis ﬁlterbank + Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

6 .Frequency domain perspective Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

Vector space interpretation   Subband decomposition is the projection of the input onto a set of “analysis vectors” in the Hilbert space of square summable sequences “Analysis vector”  Consider signal in channel q n denotes shift     Synthesis ﬁlterbank is linear combination of synthesis “basis vectors” “Synthesis vector”  n denotes shift Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. 7 .

Orthonormal subband transforms   Orthonormal expansion   Analysis and synthesis vectors are identical! Vector space Convolutional  transform Filter bank Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. 8 .

since lapped orthogonal transforms with linear phase do not exist. Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. 9 .Biorthogonal transforms   Analysis vectors and synthesis vectors are not necessarily each orthogonal. but each analysis vector must be orthogonal to all but one synthesis vectors (and vice versa)     Equivalent to perfect reconstruction Important for linear-phase FIR ﬁlters.

Linear phase is desirable for images.     Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. 10 . block-wise transform   Blockwise transforms are a special case of subband decompositions with:   Number of bands m = order of transform N   Length of impulse responses of analysis/synthesis ﬁlters ≤ m Filters used in subband coders are not in general orthogonal.Subbands vs.

11 . block-wise transform (cont.) Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.Subbands vs.

12 .Frequency response of a DCT of order N=8 10 0 -10 -20 -30 0 Frequency response [dB] Frequency response [dB] 10 Frequency response [dB] Frequency response [dB] 10 Frequency response [dB] Frequency response [dB] i=0 0 0 -10 -10 -20 -20 π/2 Frequency π -30 0 i=1 π/2 Frequency π/2 Frequency π π -30 0 i=2 10 0 -10 -20 π/2 Frequency π/2 Frequency π π -30 0 i=3 π/2 Frequency π/2 Frequency π π 10 0 -10 -20 -30 0 Frequency response [dB] Frequency response [dB] Frequency response [dB] Frequency response [dB] 10 Frequency response [dB] Frequency response [dB] 10 0 0 Frequency response [dB] Frequency response [dB] i=4 π -10 -20 π/2 Frequency π/2 Frequency π -30 0 i=5 -10 -20 π/2 Frequency π/2 Frequency π π -30 0 i=6 10 0 -10 -20 π/2 Frequency π/2 Frequency π π -30 0 i=7 π π/2 Frequency π/2 Frequency π Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

13 .Lapped Orthogonal Transform   Orthonormal convolutional transform with perfect reconstruction    Lapped orthogonal transform (LOT): only A[0] and A[1] nonzero. hence Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

Levinsky. m=8 [G. EE392C class project. 1997] Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. 14 .Example LOT basis functions.

426 bpp DCT quantizer step size 70 entropy 0. 1997] Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.LOT vs. Levinsky. EE392C class project.453 bpp [G. 15 . DCT coding LOT quantizer step size 70 entropy 0.

16 .Discrete Wavelet Transform   Recursive application of a two-band ﬁlter bank to the lowpass band of the previous stage yields octave band splitting: frequency   Same concept can be derived from wavelet theory: Discrete Wavelet Transform (DWT) Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

17 .Cascaded analysis / synthesis ﬁlterbanks Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

2-d Discrete Wavelet Transform Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. 18 .

2-d Discrete Wavelet Transform example Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. 19 .

20 .2-d Discrete Wavelet Transform example Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

21 .2-d Discrete Wavelet Transform example Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

22 .2-d Discrete Wavelet Transform example Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

23 .2-d Discrete Wavelet Transform example Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

Two-channel ﬁlterbank 2 2 2 2 ∑ Aliasing   Aliasing cancellation if : Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. 24 .

Tabatabai.Example: two-channel ﬁlter bank with perfect reconstruction   Impulse responses. synthesis ﬁlters Lowpass Highpass 1 ( ) h1 e jω ( ) “Biorthogonal 5/3 ﬁlters” “LeGall-Tabatabai ﬁlters” [LeGall. analysis ﬁlters: Lowpass Highpass     Mandatory in JPEG2000 Frequency responses: frequency response h0 e jω ( ) g1 e jω ( ) 2 g0 e jω   Impulse responses. 25 . 1988] 0 0 π 2 frequency ω π Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

26 .Quadrature Mirror Filters (QMF)   QMFs achieve aliasing cancellation by choosing Example: 16-tap QMF ﬁlterbank [Croisier. Esteban. 1976]     Highpass band is the mirror image of the lowpass band in the frequency domain Need to design only one prototype ﬁlter Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. Galand.

1986]   Impulse responses   Perfect reconstruction: ﬁnd power complementary prototype ﬁlter   Orthonormal subband transform! Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.Conjugate quadrature ﬁlters   Achieve aliasing cancelation by Prototype ﬁlter [Smith. 27 . Barnwell.

28 .Lifting   Analysis ﬁlters Σ Σ K0 Σ Σ K1     L “lifting steps” [Sweldens 1996] First step can be interpreted as prediction of odd samples from the even samples Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.

29 .)   Synthesis ﬁlters Σ - Σ - - Σ - Σ     Perfect reconstruction (biorthogonality) is directly built into lifting structure Powerful for both implementation and ﬁlter/wavelet design Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.Lifting (cont.

Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.Example: lifting implementation of 5/3 ﬁlters Σ 1+ z −1 4 Σ 1/2 Verify by considering response to unit impulse in even and odd input channel. 30 .

Incorporate the necessary rounding into lifting operator: Σ Σ K0 Σ   Σ K1 Used in JPEG2000 as part of 5/3 biorthogonal wavelet transform Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.Reversible subband transform     Observation: lifting operators can be nonlinear. 31 .

Wavelet compression results 0.074 bpp 0. 2001] Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.048 bpp Original 512x512 8bpp Error images enlarged [Gonzalez. Woods. 32 .

33 .  ZTR: zero-tree.  XX XX 2.  4. coefﬁcient and all descendants are not signiﬁcant IZ: coefﬁcient is not signiﬁcant. ZTR and IZ symbols are merged into one symbol Z Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no.Embedded zero-tree wavelet algorithm X XX XX X X X X X X X X X X X X XXXX   X X „Parent“ XX XX   „Children“   Idea: Conditional coding of all descendants (incl. children) Coefﬁcient magnitude > threshold: signiﬁcant coefﬁcients Four cases 1. but some descendants are signiﬁcant POS: positive signiﬁcant NEG: negative signiﬁcant   For the highest bands.  X X X X X X X X X X X X XXXX X X X X X X X X X X X X XXXX 3.

POS. NEG   Repeat subordinate and dominant passes.   Repeat dominant pass •  Omit previously found signiﬁcant coefﬁcients •  Decrease threshold θ by factor of 2. POS. determine new signiﬁcant coefﬁcients •  Arithmetic coding of symbols ZTR.)   Successive approximation quantization and encoding   Initial „dominant“ pass •  Set initial threshold θ.Embedded zero-tree wavelet algorithm (cont. NEG   Subordinate pass •  Reﬁne magnitude of all coefﬁcients found signiﬁcant so far by one bit (subdivide magnitude bin by two) •  Arithmetic coding of sequence of zeros and ones. Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. until bit budget is exhausted. IZ. 34 . determine signiﬁcant coefﬁcients •  Arithmetic coding of symbols ZTR. IZ.

M.Embedded zero-tree wavelet algorithm (cont. no. Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. Pearlman. 243-250. Said. December 1993. „A new. vol. fast. pp. 12. 3445-3462. 41.)       Decoding: bitstream can be truncated to yield a coarser approximation: „embedded“ representation Further details: J. pp. A. „Embedded image coding using zerotrees of wavelet coefﬁcients. 63 . vol. June 1996.“ IEEE Transactions on Signal Processing. and efﬁcient image codec based on set partitioning in hierarchical trees. A. Enhancement SPIHT coder: A.“ IEEE Transactions on Circuits and Systems for Video Technology.. Shapiro. 35 . W.

lossless) wavelet transform Zero-trees: exploit statistical dependencies across subbands Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. 36 .Conclusions:  subband and wavelet coding                 Overlapping basis functions can reduce blocking artifacts Biorthogonal subband transforms = perfect reconstruction DCT can be understood as a (poor) ﬁlter bank Discrete Wavelet Transform = cascaded dyadic subband splits Quadrature mirror ﬁlters and conjugate quadrature ﬁlters: aliasing cancellation Lifting: powerful for implementation and wavelet construction Lifting allows reversible (i..e.

Marcellin. Le Gall. 41.2.” IEEE Transactions on Acoustics.Reading         Taubman. Shapiro. Tabatabai. 37 . pp. and Signal Processing.” Proc. “Embedded image coding using zerotrees of wavelet coefﬁcients. no. 553-559. 761-764. April 1989. Speech. D. 6. Bernd Girod: EE398A Image and Video Compression Subband and Wavelet Coding no. pp. Staelin. Malvar. M.4. 12. Dec. 2. S. 7 H. vol.” IEEE Transactions on Signal Processing. pp. J. vol. D. 3445-3462. “The LOT: transform coding without blocking effects. Sections 4. H. no. 4. “Sub-band coding of digital images using symmetric short kernel ﬁlters and arithmetic coding techniques. April 1988. 37. A. 1993.1-6. ICASSP-88.” vol.

scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->