Computer Science · Computer Vision and Pattern Recognition
Biorthogonal Wavelets for Image Compression
See on the mapAbstract
Biorthogonal wavelets or filterbanks are shown to be superior in coding gain performance to orthogonal ones for logarithmic subband decompositions (limited to iterative decomposition of the downsampled output of the analysis low-pass filter). As a consequence, for logarithmic decompositions, the optimal filter is not an ideal filter. This is shown for maximally regular biorthogonal and orthogonal filters, as well as filters designed to optimize the subband coding gain. Keywords: data compression, coding gain, biorthogonal, linear phase, optimization, maximally regular 1. INTRODUCTION Because of the high theoretical performance of subband coding with ideal filters, the design of orthogonal filters with a very large stopband attenuation has been of prime concern to researchers in subband image coding. An active research area in subband coding is known as tree-structured subband coding, for which an M-band filterbank can be decomposed as several iterations of identical two-channel filterb...