An Optimal Tchebichef Moment Quantization Using Psychovisual Threshold for Image Compression

摘要

Colour image carries a certain amount of perceptual redundancy for the human eyes. The human eye is capable of perceiving various levels of colours. The sensitivity of human eye is useful for perceptual visual quality image in image compression. The visual sensitivity of the colour image in terms of image compression can be measured by a psychovisual threshold to generate the quantization tables. This paper will investigate a psychovisual threshold level for Tchebichef moment transform (TMT) from the contribution of its moments. In this paper presents a new technique to generate quantization table for TMT image compression based on psychovisual error threshold. The experimental results show that these new finer quantization tables based on psychovisual threshold for TMT provides better quality image and lower average bit length of Huffman code than previously proposed TMT quantization.