Vector Quantization Based on Visual Sensitivity to Image Edges

摘要

Vector quantization (VQ) has received considerable attention and becomes an effective tool for image compression. It provides high compression ratios and simple decoding processes. However, studies on practical implementation of VQ have revealed some major difficulties such as edge distortion and codebook design efficiency. In order to solve them, this paper introduced a new VQ method by self-organizing feature map (SOFM). Mainly, two kinds of weighting factor are proposed, which are based on the complexity of images and the feature of image edges respectively. The adaptive learning procedure is performed using weighting factor. As a result, an implication of the proposed algorithm to image compression problem gives improved edge characteristics in reconstructed images. In comparison, the reconstructed images using the latter weighting factor have better visual quality than using the former.