Application of HADAMARD Transform For Fast VQ Image Coding

Abstract

Image compression using vector quantization has been proved to be an efficient approach. The speed of codeword search is the key problem for real-time applications of vector quantization. In this paper, a new inequality based on the HADAMARD transform is derived. An efficient VQ codeword search algorithm using this inequality is proposed for VQ image coding. Experimental results show that the algorithm performs better than the PDS, ENNS, IENNS, and WTPDS algorithms.