Linear-phase perfect reconstruction filter bank: lattice structure, design, and application in image coding

摘要: A lattice structure for an M-channel linear-phase perfect reconstruction filter bank (LPPRFB) based on the singular value decomposition (SVD) is introduced. The can be proven to use a minimal number of delay elements and completely span large class LPPRFBs: all analysis synthesis filters have same FIR length, sharing center symmetry. also structurally enforces both properties, capable providing fast efficient implementation, avoids costly matrix inversion problem in optimization process. From block transform perspective, new viewed as representing family generalized lapped biorthogonal (GLBT) with arbitrary channels M arbitrarily overlap. relaxation orthogonal constraint allows GLBT significantly different basis functions, which then tailored appropriately fit particular application. Several design examples are presented along high-performance GLBT-based progressive image coder demonstrate potential transforms.