周 强,王志强,杨贵琳,李清华.基于压缩感知技术的纸病图像数据实时采集研究[J].中国造纸学报,2015,30(3):51-56 本文二维码信息
二维码(扫一下试试看!)
基于压缩感知技术的纸病图像数据实时采集研究
Study of the Real-time Acquisition and Transmission of Paper Disease Image Based on Compressed Sensing
  
DOI:10.11981/j.issn.1000-6842.2015.03.51
中文关键词:  压缩感知  纸病图像  数据实时采集和传输
Key Words:compressed sensing  paper disease image  data real-time acquisition and transmission
基金项目:陕西省科技统筹创新工程计划项目(2012KTCQ01-19);陕西省科技攻关项目(2011K06-06);西安市未央区科技计划项目201304。
作者单位
周 强1 1.陕西科技大学电气与信息工程学院, 陕西西安710021 
王志强1,* 1.陕西科技大学电气与信息工程学院, 陕西西安710021 
杨贵琳1 1.陕西科技大学电气与信息工程学院, 陕西西安710021 
李清华2 2.西安西翰电力科技有限公司陕西西安710065 
摘要点击次数: 4294
全文下载次数: 1192
中文摘要:
      压缩感知理论基于信号的稀疏性,压缩感知技术在采集信号的同时,实现数据的压缩处理,能够显著减少传输过程中纸病图像的数据量。结合纸病图像的特点,在研究了纸病图像稀疏性的基础上确定了测量矩阵,完成了计算机PC重建时的重构算法。通过仿真实验,验证了不同的稀疏基和采样率对纸病图像重构质量的影响。结果表明,利用压缩感知技术,纸病图像数据的传输量只有原来的30%~40%,并且重构的图像质量也较好,能够在一定程度上提高造纸生产线上纸病检测的速度。
Abstract:
      With the increase of machine’s speed and paper width in the paper industry, the amount of image data acquisition and transmission is larger, the poor real-time of on-line detection has been a bottleneck in the paper disease detection. Based on sparsity of the signal, compressed sensing can realize the compression of the data while the data are collecting, and significantly reduce the data amount in the transmission. Combining with the characteristics of paper disease images, this paper determined the measurement matrix based on the study of the sparsity, and obtained reconstruction algorithm on PC. Through the simulation experiment, the impact of different sparse matrix and different sampling rate on the quality of the disease image reconstruction was verified. The result showed that, using this technology, data transmission amount reduced 60%~70% and the quality of the reconstructed image was good, this could improve the speed of paper disease detection to a certain extent.
查看全文  查看/发表评论  下载PDF阅读器  HTML

分享按钮