A New Implementation of Particle Filter for Digital Noisy Images

This paper proposes a new method Rao-Blackwellized Particle Filter with Maximum Likelihood Estimation for reducing noise in digital images. The proposed method first estimates the noise level of the real noisy image using simple and fast algorithm and then the method is used for noise reduction. This method computes a highly accurate proposal distribution based on the maximum likelihood observations. Extensive experimental results demonstrate that our method can obtain better performances in terms of both subjective and objective evaluations than the state-of-the-art denoising techniques.