Multigrade compression algorithm for image based on better frequency resolution of wavelet transform

This paper presents several efficent experimental systerms of compression for chosing quality of image restoration. By analyzing and chosing of wavelet transform for every frequency band carefully, We proposal five compressional grades based on effect of human visual system. That is super-fine,fine,general, admissible and coarse grained. Their objective quality of restoration are 38db,37db,36db and 35db. Theoretically,the original image is decomposed at different scale or resoluion levels onto the orthogonal wavelet basis which has one or half octave,then coded using multiscale vector quatization techniques.TelecommunicationsCPCI-S(ISTP)