图像处理和重构算法是单像素成像系统中重要环节.本研究针对QR码的重建,通过对比7种常用的传统图像处理和重构算法,研究图像重建效果,其中全变分压缩感知算法(Total Variation Algorithm,TV)成像效果最优.最后采用TV算法和基于TV算法优化的TVAL3算法(Total Variation Augmented Lagrangian Alternating Direction Algorithm)分别对QR码和常见的图片进行重建试验.结果显示,TVAL3算法的重建速度和效果优于TV算法,无论对简单图像或是复杂图像均表现出较好的成像效果.本研究可为单像素成像图像重建算法应用提供有价值的参考.
Research on Image Reconstruction Algorithms Based on Single Pixel Imaging
Image processing and reconstruction algorithms are important links in single pixel imaging systems.In this study,focused on the reconstruction of QR codes,compared 7 commonly used traditional image processing and reconstruction algorithms,the image reconstruction effects were studied.Among them,the total variation (TV) algorithm has the best imaging effect.Finally,TV algorithm and TVAL3 algorithm based on TV algorithm optimization were used to reconstruct QR codes and common images,respectively.The results showed that TVAL3 algorithm had better reconstruction speed and effect than TV algorithm,and obtained good imaging performance for both simple and complex images.This study can provide a useful reference for the application of single pixel imaging reconstruction algorithms.
Single pixel imagingImage reconstruction algorithmCompression perceptionSparse expressionTVAL3
NETL
NSTL
万方数据
