MULTIPLE BLOCK-SIZE TRANSFORM VIDEO CODING USING AN ASYMMETRIC SUB-BAND STRUCTURE

A sub-band coder for video signals is disclosed, which is easily realizable in VLSI due to its simple structure. The coder uses sets of two-tap finite impulse response (FIR) filters having equal-magnitude coefficients to decompose the video signal at multiple stages. In the preferred embodiment the sub-band coder has an asymmetrical structure. With the arrangement, the sub-band decomposition effects what is equivalent to a multiple block-size Hadamard transform of the video signal in which coefficients are transmitted from various block-sized transforms of larger blocks of pixel data.