Arabic Textual Images Compression Approach

In this paper, a novel compression approach for large Arabic textual images is proposed. Initially, input texts are segmented into patterns which represent sub-words. Then, patterns-matching procedure is used in order to find similar patterns within the image. Finally, to optimize the performances of this approach, an adaptive arithmetic coding is used to encode the resulting data streams. Experimental results show that the average compression ratio of our approach is better than other existing algorithms.