A Novel FPCM Based Vector Quantizer Codebook Design for Image Compression in the Wavelet Packet Domain

This paper presents a novel Vector Quantizer codebook design using Fuzzy Possibilistic C Means (FPCM) clustering technique for image compression using Wavelet Packet. The idea is to achieve higher compression ratio based on clustering the wavelet coefficients of each Wavelet Packet Tree (WPT) bands. The methodology applied here is to apply WPT to the whole image. The sub blocks are decomposed into two level Wavelet Packet Tree where the coefficients of LL band’s(approximation and details), LH band’s(approximation), HL band’s(approximation), HH band’s (approximation) are clustered using FPCM. The centroids of each cluster is arranged in the form of a codebook and indexed. The index values are coded and then transmitted across. The image is reconstructed using the inverse WPT followed by rearranging and the subsequent encoder. The results show that the psycho-visual fidelity criteria (both subjective and objective measures) of the proposed FPCM technique do better than the other existing methods.