摘要

In this paper, a novel wavelet image coding algorithm based on human visual system characteristics is proposed. After dividing the original image into sub-blocks, including smooth block, edge block and texture block, different perceptual weights and compression algorithms are applied to each sub-block. Optimal coding of the whole image is achieved by choosing perceptual weights and compression algorithms according to the human visual system characteristics of each sub-block. An innovated image coding process, in which different signal transforms are combined together to provide an efficient and effective image representation is put forward here. Compared with traditional single wavelet image coding algorithm, substantially improvements of visual quality of reconstructed image and coding performance speed are shown in experimental results.