Low Bit Rate Analysis of Adaptive Synthesis Filter Banks for Image Compression
Abdul Nadeem, Yangyu Fan · International Journal of Hybrid Information Technology · 2014
The aim of this paper is to analyze the performance of adaptive synthesis filter banks for image compression at low bit rates.In adaptive synthesis filter banks, analysis section comprises of linear phase filters whereas, synthesis section manages a combination of linear phase and non linear filters (in form of delay filters).The adaptive synthesis filter banks have advantages over conventional adaptive filter banks because they have no synchronization issues between analysis and synthesis filters and are compatible with the existing subband coding systems.Their performance is evaluated against various compression ratios by using biorthogonal filters '4.4' and '5.5' which suggest adaptive synthesis filter banks achieve significant gain in PSNR at compression ratio 0.3 ~ 0.6 bit per pixel.