基于压缩感知的地磁数据分频压缩与重构方法.pdfVIP

;uJ/1 拎揣 第 42 卷第4 期 训 ，瓦 二子 m Vo l. 42 , NO.4 2013 年 8 月 Acla Geodaetica et Cartographica Si r1 lca Aug. ,2013 LIU Fanming , ZHANG Yingfa , LI Ya口， et al. Geomagnetic Data Frequency Divisive Compression and Reconstructíon Based on Compressed Sensing[J]. Acta Geodaetica et Cartographica Sinica ,2013 ,42(4) :481-486. (刘繁明，张迎发，李艳，等.基于压缩感知的地磁 数据分频压缩与重构方法口]. ÌÝ!~绘学报， 2013 ， 42(4):481-486.) 基于压缩感知的地磁数据分频压缩与重构方法 刘繁明，张迎友，李 艳，姚建奇 哈尔滨工程大学自动化学院，黑龙江哈尔滨 150001 Geomagnetic Data Frequency Divisive Compression and Reconstruction Based on Compressed Sensing Ll U Fanming , ZHANG Yingfa , LI Yan , YAO Jianqi College 01 Automation , Harbin Engineering University , Harbin 150001, China Abstract :It is necessary to compress the geomagnetic data in order to establish the reference database for naviga- \ion. In order \0 reserve \he detai Is of high frequency which contai 口 the texture or edge feature of the image , a novel method for geomagnetic data compression and reconstruction is presented. Algorithm is based upon discrete cosine transform for data frequency divisive compression , and takes advantage of compressed sensing. The accurate reconstruction is performed by partial cosine transform measurements , and the methods consis\ of solving lj-nOrm a 门 d total-variation minimization programs. The simulation based on actual measured geomagnetic data shows that , faithful geomagnetic data can be reconstructed from a subset that represents a mere 25 percent of the complete set of measurements. Comparing with traditio门al filtered back projection methods , it gives better apparent advantages in reconstruction quality and robustness , and the most increase value of peak signal to no/se ratio is 5 dB. Key words :navigation database;