Abstract
Digital image compression technology is of special interest for the fast transmission and real-time processing of digital image information. Image compression is a technique developed for a long time, plus there are several approaches which reduce the compression rate, and accelerate computation time, still we need a lot to go to improve the efficacy of compression. In this paper, we are compressing document files containing images and text. In comparisons to state-of-the-art commercial MRC products and selected segmentation algorithms in the literature, our new algorithm achieves greater accuracy of non-text detection but with a lower false detection rate of text features.
