基于OpenMP的压缩感知多描述并行处理算法
巫小婷;邓家先;任玉莉;杨扬
【期刊名称】《计算机应用研究》 【年(卷),期】2013(030)004
【摘要】Concerning the long time running for reconstruction and the poor quality of reconstruction image in compressed sensing, this paper proposed a multiple description of compressed sensing parallel processing algorithm on the basis of a careful analysis. In order to improve the speed and the quality of reconstruction image, divided the coefficients after sparse transform into several sub-images using interleaving extraction. And then assigned these sub-images to each thread to using OpenMP. The experimental results show that the quality of reconstruction image is rising with the increase of the extraction number, compare with the single extraction, the PSNR in 32-extraction is higher 7. 24 dB. Moreover the executed efficiency continues to increase with the increase of the number of threads, it' s up to 7. 35 times compare with the single-thread. It solved the problems of long time running and poor quality of reconstruction image effectively by the proposed algorithm.%针对压缩感知算法重建时间长、图像重建质量不高等不足,在认真分析压缩感知算法的基础上,提出一种压缩感知多描述并行算法.为了提高系统运行速度和重建图像质量,将经过稀疏变换后的系数进行交织抽取,分成多个子图像,再利用OpenMP将子图像分配到各线程中并行实现分块压缩
感知.实验结果表明,随着抽取数的增加,图像的重建质量呈上升趋势,在32-抽取时图像的重建质量比单抽取的高出了7.24 dB;随着线程数的增加,程序的执行效率不断提高,最高可达单线程的7.35倍,有效地解决了压缩感知算法重建时间长以及图像重建质量差的问题. 【总页数】3页(1278-1280)
【关键词】压缩感知;多描述;OpenMP;并行;交织抽取;加速比 【作者】巫小婷;邓家先;任玉莉;杨扬
【作者单位】海南大学信息科学技术学院,海口570228 【正文语种】中文
【中图分类】TP301.6;TP391.4;TN911.73 【文献来源】
https://www.zhangqiaokeyan.com/academic-journal-cn_application-research-computers_thesis/0201241978586.html 【相关文献】
1.基于OpenMP的压缩感知并行处理算法 [J], 巫小婷; 邓家先
2.基于海量物联网数据的压缩感知及其并行处理 [C], ZHOU Jian; 周健; ZHANG Yong-ping; 张永平; ZHANG Gong-xuan; 张功萱 3.一种基于压缩感知的快速图像处理算法研究 [J], 刘昆
4.无线传感器网络基于压缩感知数据处理算法研究 [J], 庄严; 沈军; 庄田 5.一种基于压缩感知的票据图像处理算法 [J], 梁添才; 聂芸芸; 王琨; 张永
以上内容为文献基本信息，获取文献全文请下载
基于OpenMP的压缩感知多描述并行处理算法
