Rank-based image transformation for entropy coding efficiently

In this paper, we introduce a rank-based image transformation which is pre-processing method for gray-level images to be compressed more efficiently by entropy encoder. Before entropy encoding a stream of gray-level values in an image, the proposed method counts co-occurrence frequencies for neighboring pixel values. Then, it replaces each pairs of adjacent gray values with particular ordered number based on the investigated co-occurrence frequencies. Finally, the method transmits the adjusted data which has ordered number to entropy encoder. Because statistical characteristic is more enhanced by this preprocessing step, it is possible to improve performance of entropy coding. From the simulation result using 8 bits gray-scale images, it is verified that the proposed method can reduce bit rate by up to 37.85% than plain entropy coders.