Traditional second-order correlation reconstruction method required a large number of measurements, in which not only the quality of reconstructed image was poor but also didn't meet the real-time requirements. We combine the total variation with the compressive sensing method to reconstruct the object image in ghost imaging. The paper describes the basic structure of objective function based on total variation regularization and the corresponding compressive sensing recovery algorithm, and take a comparison with the gradient projection based compressive sensing algorithm about the recovery performance. The simulation results show that compressed sensing algorithm based on total variation regularization has a better compared reconstruction performance than algorithm based on gradient projection algorithm in ghost imaging system. Then apply the above algorithms to experimental data of ghost imaging experiment, and finally got the reconstructed images of the target image. The results once again demonstrate the effectiveness and feasibility of the algorithm based on total variation.

Date Published: 5 January 2017
PDF: 6 pages
Proc. SPIE 10244, International Conference on Optoelectronics and Microelectronics Technology and Application, 102440X (5 January 2017); doi: 10.1117/12.2264429

Published in SPIE Proceedings Vol. 10244:
International Conference on Optoelectronics and Microelectronics Technology and Application
Shaohua Yu; Jose Capmany; Yi Luo; Yikai Su; Songlin Zhuang; Yue Hao; Akihiko Yoshikawa; Chongjin Xie; Yoshiaki Nakano, Editor(s)