BWT-based Data Preprocessing for LZW

In this paper we propose a BWT-based LZW algorithm for reducing the compressed size and the compression time. BWT and MTF can expose potential redundancies in a given input and then significantly improve the compression ratio of LZW. In order to avoid the poor matching speed of LZW on long runs of the same character, we propose a variant of RLE named RLE-N. RLE-N does not affect the compression ratio, but it contributes LZW to reduce the execution time obviously. The experimental results show that our algorithm performs well on normal files.

[1]  Raffaele Giancarlo,et al.  Boosting textual compression in optimal linear time , 2005, JACM.

[2]  Bernhard Balkenhol,et al.  Modifications of the Burrows and Wheeler data compression algorithm , 1999, Proceedings DCC'99 Data Compression Conference (Cat. No. PR00096).

[3]  Yao-hua Xie,et al.  Image Compression Based on Classification Row by Row and LZW Encoding , 2008, 2008 Congress on Image and Signal Processing.

[4]  Terry A. Welch,et al.  A Technique for High-Performance Data Compression , 1984, Computer.

[5]  Michael G. Strintzis,et al.  A context based adaptive arithmetic coding technique for lossless image compression , 1999, IEEE Signal Processing Letters.

[6]  D. J. Wheeler,et al.  A Block-sorting Lossless Data Compression Algorithm , 1994 .

[7]  Chin-Chen Chang,et al.  High-Capacity Reversible Data-Hiding for LZW Codes , 2010, 2010 Second International Conference on Computer Modeling and Simulation.

[8]  Rahul Singh,et al.  Influence of Dictionary Size on the Lossless Compression of Microarray Images , 2007, Twentieth IEEE International Symposium on Computer-Based Medical Systems (CBMS'07).