High-quality correspondence imaging based on sorting and compressive sensing technique

We propose a high-quality imaging method based on correspondence imaging (CI) using a sorting and compressive sensing (CS) technique. Unlike the traditional CI, the positive and negative (PN) subsets are created by a sorting method, and the image of an object is then recovered from the PN subsets using a CS technique. We compare the performance of the proposed method with different ghost imaging (GI) algorithms using the data from a single-detector computational GI system. The results demonstrate that our method enjoys excellent imaging and anti-interference capabilities, and can further reduce the measurement numbers compared with the direct use of CS in GI.