Comparing Vector Quantization and Wavelet Coefficients

690 Words2 Pages

Wavelet transform is efficient tool for image compression,
Wavelet transform gives multiresolution image decomposition. which can .be exploited through vector quantization to achieve high compression ratio. For vector quantization of wavelet coefficients vectors are formed by either coefficients at same level, different location or different level, same location.
This paper compares the two methods and shows that because of wavelet properties, vector quantization can still improve compression results by coding only important vectors for reconstruction. Thus giving priority to the important vectors higher compression can be achieved at better quality. The algorithm is also useful for embedded vector quantization coding of wavelet coefficients.
INTRODUCTION
lmage compression reduces the data for image representation. lmage compression techniques are divided into two main categories Transform techniques and nontransform techniquesh transform techniques compression is achieved by encoding the transformed coefficients.
Compression based on wavelet transform decomposes the image in four subbands. These subbands represent the image at coarse and fine resolution in different orientations. [l]
For lossy encoding of coefficients scalar or vector quantization can be used. In vector quantization group of coefficients are Considered instead of individual pixels.
Closest match is found for the given code vector id found from codebook using some criterion. This gives higher compression ratio. Different vector quantization methods arc suggested for wavelet coefficients [2]. For all these methods vector formation is considered by two ways.
I Intraband-, which groups the pixel at same level and same location. 2. Inter-band which groups the pixels...

... middle of paper ...

...( 1996)
Jason Knipe, Xiaobo Li, Bin Han ‘‘ An improved lattice vector quantization scheme for wavelet compression” IEEE transaction on signal processing
Vol. 46 no. Ipp 239-243 January I998
,,
4
5.
6.
Young Huh, 1.1 Hwang , and K.R.Rao “Block wavelet transform coding of images using classified vector quantization” IEEE transaction on circuits and systems for video technology vo1.5 no.1 February
1995 pp. 63-67
Amir Averbuch, Danny Lrar Moshe Israeli Image compression using wavelet transform and multirebolution decomposition” IEEE transaction on image processing vo1.5 no.1 January 1996 pp. 4-12
Madhuri Khambete, Dr. Madhuri Joshi ’‘ Adaptive vector quantization based on quality criterion using
Hosaka plot” tencon99 vol.. pp. - 754-756
Original boat image METHOD1
PSNR = 22.47 db
Cnmnrecsion cltin 77 bird nixd
PROPOSED METHOD
PSNR= 25.23 db
Compression ratio .22 blpixel

More about Comparing Vector Quantization and Wavelet Coefficients

Open Document