Iterative first-order reverse image filtering

Li Dong, Jiantao Zhou, Cuiming Zou, Yulong Wang

Abstract: In this paper, we study an unconventional but practically meaningful problem reverse image filtering, which aims at removing the filtering effects with the given image filter. We propose an iterative first-order reverse image filtering algorithm based on the Newton-Raphson method. The convergence of the proposed iterative procedure is verified via extensive experiments. The experimental results show that our method could effectively reverse many prevalent image filters. Our method finds its practical usage in many applications, such as non-blind image deblurring and image details recovery. Furthermore, our method could probe the behaviors of black-box image filtering systems, which poses security concerns on the protection for the privately commercial image filters.