De-noising ghost imaging via principal components analysis and compandor

Improving the ghost imaging quality and speed remains a challenging task. Here, we propose an optimization algorithm model to the ghost imaging by employing principal components analysis and companding technique. By choosing appropriate parameters of the principal components and the companding function, the ghost image quality is enhanced. A good agreement between the simulation and the experiment result is obtained. In addition, we demonstrate the method with a complicated sample compared with the other five existing algorithms, indicating its advantages for wide range of applications. At last, a criteria function is firstly proposed and built to optimize the parameters for better reconstruction result without the prior information of the object. This optimization model may offer a promising implementation of de-noising ghost imaging.