基于SPIHT编码的管道泄漏检测数据压缩方法  

Compression algorithm of pipeline leak detected data based on SPIHT coding

作　　者：刘琳琳[1] 王琳[1] 王学伟[1] 

机构地区：[1]北京化工大学信息科学与技术学院,北京100029

出　　处：《辽宁工程技术大学学报（自然科学版）》2011年第2期280-283,共4页Journal of Liaoning Technical University (Natural Science)

基　　金：国家自然科学基金资助项目(50975019)

摘　　要：针对油气管道泄漏检测数据的压缩问题,首次提出基于小波变换的多级树集合分裂编码(SPIHT)的泄漏检测数据压缩方法。该方法主要是利用原始信号在各尺度下小波变换系数的自相似性,优先传送绝对值较大的小波系数,再传送绝对值相对较小的作为非重要系数以得到更高的压缩倍数和更小的重构误差。应用该方法对实际泄漏检测数据进行压缩,实验结果表明:在不大于1%的均方根误差下,压缩前和压缩后数据的压缩倍数相差达到28,比用EZW算法和算术编码相结合的方法提高了12倍,更多地节省了泄漏检测数据的存储空间。To improve the data compression efficiency on oil and gas pipeline leak detection data,a new method of data compression based on wavelet transform and Set Partitioned in Hierarchical Tree（SPIHT） coding is presented in this paper.The method mainly uses the self-similarity of wavelet transform coefficient of original signal under various scales.In order to achieve a higher compression ratio and a smaller reconstruction error,the method requires the transmission of the wavelet coefficients with larger absolute values first,and then the transmission of those with the smaller values.The experimental results demonstrate that the compression ratio is 28 when the ratio of signal and noise is less than or equal to 1%.The compression ratio increases 12 times than that using the method which combines the EZW algorithm and the arithmetic coding.The proposed method can save more storage space for leakage detection data.

关 键 词：输油管道 管道泄漏检测 数据压缩 小波变换 多级树集合分裂算法 

分 类 号：TM76[电气工程—电力系统及自动化]