Guided Quaternary Reaching Method for Wavelet-based Image Compression

Summary form only given. In this paper, a simple - easy to understand, easy to implement - high performance wavelet-based image compression algorithm using guided quaternary reaching (GQR) method is proposed. Compared with JPEG2000, GQR algorithm has two remarkable merits: 1). much lower algorithmic complexity and 2). higher compression efficiency In a wavelet-based image compression system, wavelet coefficients are rearranged into hierarchical subbands. By nature, image signals are 2-dimensional and hence the significance of wavelet coefficients are 2-dimensionally distributed within each subband. After quantization, most coefficients become small integers or zeros. To a relatively large threshold, the whole scenery sees that significant coefficients sparsely and 2-dimensionally disperse in a wide tract of insignificant coefficients. Therefore, linear scan patterns inevitably cause losses in compression efficiency.