摘要
采集可压缩信号时,单像素相机的测量矩阵和稀疏变换基共同构成重构矩阵。针对0-1循环测量矩阵,尽管利于编程和硬件实现,但相应重构矩阵重构能力很差的问题,对比分析了基于0-1随机矩阵和0-1循环矩阵的重构矩阵,及其优化前后的重构效果,并给出相应的理论分析和比较评估。该理论分析和比较评估为单像素相机测量矩阵的选型和设计提供了方向指导,利于设计制造出性能更优的单像素相机。实验表明,对于可压缩信号,单像素相机不能使用0-1循环矩阵作为测量矩阵,但可使用0-1随机矩阵作为测量矩阵,并能对相应重构矩阵作有效优化。
When a single-pixel camera collecting compressible signals,the reconstruction matrix is formed by the measurement matrix and the sparse transform base.Although the 0-1 circulant measurement matrix is conducive to programming and hardware implementation,the reconstruction ability of the corresponding reconstruction matrix is very poor.The reconstruction matrices based on the 0-1 random matrix and the 0-1 circulant matrix and reconstruction effects before and after the optimization were compared and analyzed.And corresponding theoretical analysis and comparative evaluation were given.The theoretical analysis and comparative evaluation could guide the selection and design of the measurement matrix of the single-pixel camera,and facilitated the design and manufacture of a single-pixel camera with better performance.Experiment results showed that,for compressible signals,a single-pixel camera could not use the 0-1 circulant matrix as the measurement matrix;however,the 0-1 random matrix could be used as the measurement matrix,and the corresponding reconstruction matrix could be effectively optimized.
作者
程涛
吴小龙
杨明
CHENG Tao;WU Xiaolong;YANG Ming(School of Mechanical and Transportation Engineering, Guangxi University of Science and Technology, Liuzhou 545006, China)
出处
《探测与控制学报》
CSCD
北大核心
2021年第2期81-85,共5页
Journal of Detection & Control
基金
国家自然科学基金项目资助(41461082,81660296)
中国博士后科学基金项目资助(2016M592525)
广西自然科学基金项目资助(2014GXNSFAA118285)
广西高校科学技术研究项目资助(YB2014212)
广西科技大学博士基金项目资助(校科博13Z12)。
关键词
单像素相机
离散余弦变换
测量矩阵
数字微镜
0-1循环矩阵
single pixel camera
discrete cosine transform(DCT)
measurement matrix
digital micromirror device
0-1 circulant matrix
