WAVELET IMAGE ENCODING AND DECODING DEVICE AND METHOD USING HUMAN VISUAL SYSTEM MODELING

摘要

Input image data is wavelet-transformed into wavelet transformed data. First, second and third quantization values respectively corresponding to the frequency variation, background brightness and edge state of the wavelet transformed data are determined. The wavelet transformed data is quantized by using a human visual quantization value for each region corresponding to the first, second and third quantization values. The quantized data is entropy-encoded to generate compressed image data.