Indian Journal of Science and Technology

Abstract

The Communication of Scanned document needs high bandwidth due to the high volume of data and large size of the document. In order to optimize the bandwidth in uploading, downloading and exchanging the operation in multimedia communication, compression techniques are essential. Many compression techniques are already available in the literature such as Multi-dimensional Multi-scale Parser (MMP), JPEG-2000, Digipaper, DjVu, and H.264/AVC. However the compression ratio and PSNR values provided by them are not sufficient to handle the large volume of documents. Hence a new compression algorithm called enhanced-MMP is proposed in this paper by enhancing the existing MMP, R-D segmentation and optimization techniques have been used in this work and the compression algorithm uses quad tree based segmentation for smooth image and binary tree based segmentation for non-smooth image. This greatly reduces the complexity of segmentation and finally an adaptive dictionary matching technique is used for matching the pattern effectively, which contains different scale dimensions. This algorithm provides better performance than the existing algorithms with respect to compression ratio and PSNR.
Keywords: Dictionary Revision Method, Pattern Matching and EMMP, Predictive Coding, Quad Tree Segmentation, R-D Segmentation, R-D Optimization