2D Compressed Sensing Using Nonlocal Low-Rank Prior Reconstruction for Cipher-Image Coding

In recent years, cipher-image coding by using compressed sensing (CS) has became a hot topic. However, the ratio-distortion (R-D) performance of the previous methods are barely satisfactory. In order to address this concern, a 2D CS (2DCS) scheme by using nonlocal low-rank prior (NLP) reconstruction is proposed in this letter. Firstly, the scrambling encryption is applied to mask the plaintext image. Secondly, the cipher image is compressed by 2DCS. Lastly, an iterative singular value thresholding (ISVT) algorithm is developed, which can reconstruct the image effectively by exploring the NLP information of the image. Simulation results show that the proposed method outperforms the previous CS-based methods in terms of R-D performance.

Automated summary

In this letter, a 2D compressed sensing (2DCS) scheme using nonlocal low-rank prior (NLP) reconstruction is proposed. By applying scrambling encryption and an iterative singular value thresholding (ISVT) algorithm, the proposed method outperforms previous CS-based methods in terms of R-D performance.