Fast non-blind deconvolution method for blurred image corrupted by poisson noise
Shuyin Tao, Wende Dong, Zhenmin Tang, Qiong Wang · 2017
We formulate the deconvolution problem by combining the negative logarithmic Poisson likelihood with the total variation (TV) regularization, and present a fast algorithm which is based on the method of Lagrange multiplier to solve it. In the proposed algorithm, the original problem is converted into two sub-problems. One is a simple convex optimization problem which has a closed-form solution. While the other is a conventional deconvolution problem based on the Gaussian noise model, which can be solved efficiently with the variable splitting and penalty technology. The minimizer is reached by alternately solving the two problems for only a few iterations. Experimental results show that the algorithm runs very fast and can achieve restored image of high accuracy.