New method for adaptive lossless compression of still images based on the histogram statistics

Abstract

The method for adaptive lossless image coding is aimed at efficient compression of grayscale or color digital still images. The method is based on the analysis of the processed images histograms, followed by modified Huffman and Run-length coding. The coding is performed in two consecutive stages. In the first one, the input data (i.e., the digital information about the image brightness and color components) is transformed without affecting their volume in such a way, that to obtain sequences of same values (in particular, zeros) of maximum length. The transform is reversible and is based on the analysis of the input data histograms and on the differences between each number in the processed data and the most frequent one. In the second stage of the processing the transformed data is analyzed and sequences of same numbers are detected. Every such sequence is substituted by a shorter one, corresponding to the number of same values which it contains. The method is extremely suitable for images of graphics or texts (for example: fingerprints, contour images, cartoons, medical signals, etc.). The method has low computational complexity and is suitable for real-time applications. © 2011 Springer-Verlag Berlin Heidelberg.