一种基于算术编码的文本数据压缩算法

发布时间：2018-05-18 15:31

本文选题：算术编码 + 参数优化　；参考：《电子科技大学学报》2016年06期

【摘要】：提出了一种基于算术编码的文本数据压缩算法,将扫描产生的偏移量、匹配数据长度等全局优化问题转化为局部优化问题,并从Glomb编码思路出发,推导出一种参数选择算法;对LZ77算法进行修正,提出一种预测编码方法,获得预测参数。对预测参数、偏移量、数据匹配长度、保留文本数据使用MQ算术编码器进行编码,针对不同类型数据,设计出不同的编码算法和相应的上下文算法。对算法进行仿真,并与Winzip、Win Rar压缩效率进行比较,结果表明对纯文本数据、Word文档数据、C语言程序代码,图像数据等,该压缩算法优于Winzip;在纯文本数据、Word文档数据、C语言程序代码压缩方面与Win Rar相当或者略好,但在图像压缩方面的性能与Win Rar相比略有不足。
[Abstract]:A text data compression algorithm based on arithmetic coding is proposed in this paper. The global optimization problems such as the offset generated by scanning and the length of matching data are transformed into local optimization problems. A parameter selection algorithm is derived from the idea of Glomb coding. The LZ77 algorithm is modified and a predictive coding method is proposed to obtain the prediction parameters. The prediction parameters, offset, data matching length, reserved text data are encoded by MQ arithmetic encoder, and different coding algorithms and corresponding context algorithms are designed for different types of data. The algorithm is simulated and compared with Winzip-Winin Rar compression efficiency. The results show that the pure text data, word document data, C language program code, image data, etc. The compression algorithm is superior to Winzipand is comparable to or slightly better than Win Rar in the compression of pure text data and word document data, but the performance of image compression is a little less than that of Win Rar.
【作者单位】：西安电子科技大学电子工程学院;西安电子科技大学网络行为研究中心;
【基金】：国家自然科学基金(61571354)
【分类号】：TP391.1

【相似文献】