提高关联成像重构质量的研究  

A comparative study of reconstruction methods for ghost imaging

作　　者：张雷洪[1] 张志晟 樊丽萍[1] ZHANG Leihong;ZHANG Zhisheng;FAN Liping(College of Communication and Art Design,University of Shanghai for Science and Technology,Shanghai 200092,China)

机构地区：[1]上海理工大学出版印刷及艺术设计学院,上海200092

出　　处：《光学仪器》2020年第2期32-38,共7页Optical Instruments

摘　　要：比较基于压缩感知关联成像(CGI)与伪逆关联成像(PGI)两者之间的成像效果差异,探讨形态学权重自适应对关联成像去除噪声的效果。选择不同的图像,通过MATLAB软件开展仿真实验,对目标图像分别采样64、256、512、1024、2048、3000次,首先通过关联成像、基于压缩感知关联成像与伪逆关联成像三种方法重构图像,再对比压缩感知与伪逆两种方法重构图像的效果,以峰值信噪比(PSNR)、相关系数(CC)为量化指标,将基于压缩感知关联成像与伪逆关联成像在不同采样次数下进行对比分析。同时,通过实验分析形态学权重自适应去除关联成像中噪声的效果。伪逆关联成像在低次数采样的情况下比基于压缩感知关联成像的成像效果更好,在高采样次数下,基于压缩感知关联成像的成像效果更好。在实际重构中压缩感知关联成像重构的图像仍有噪声,形态学权重自适应可以有效去除关联成像实验中产生的噪声。To compare the imaging effects between compressed sensing ghost imaging(CGI)and pseudo-reverse ghost imaging(PGI)and explore the effect of morphological weight adaptive on correlated imaging to remove noise.Different images were selected,and simulation experiments were carried out by MATLAB software.The target images were sampled with 64,256,512,1024,2048,and 3000 rate respectively.The images were reconstructed by correlation imaging,compressed sensing ghost imaging and pseudo-reverse ghost imaging.The two methods to reconstruct the image effect,and then using the peak signal-to-noise ratio(PSNR)and correlation coefficient(CC)as the quantitative indicators,the compressed sensing ghost imaging and pseudoreverse ghost imaging were used under different usage times.A comparative analysis was conducted.At the same time,in the experiment,morphological weighted adaption was used to remove noise from compressed sensing ghost imaging.pseudo-inverse ghost imaging is better than compressed sensing ghost imaging in the case of low-order sampling.Under high sampling rate,the imaging effect based on compressed sensing ghost imaging is better,but in the actual reconstruction,compressed sensing ghost imaging still has noise,and the morphological weight adaptation can effectively remove the noise in the ghost imaging experiment.

关 键 词：关联成像 目标重构 压缩感知 伪逆矩阵 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]