A high quality image compression technique for low cost multimedia applications

A novel image compression technique suitable for low cost multimedia applications is reported. The method is a class of interpolative vector quantization which, in the past, is implemented using fix-rate subsampling and interpolating processes. In the proposed technique, a new adaptive sampling point detection algorithm is employed to construct non-uniformly spaced decimation lattice. In effect, only sampling points which have significant effects on the aliasing error are extracted; hence providing superior coding fidelity at relatively lower bit-rate.<
>