In the paper, we proposed a new iterative algorithm and use a entirely new iterative factor. Firstly, we adopt the Exp function in the iterative factor, because we want each iterative result preserves the nonnegative constraint; Secondly, we make the iterative factor in a reciprocal form ,this way can produce two advantages, one is we can get a more stable and continuous results after each iteration; the other is we can achieve this algorithm in hardware more convenient. Thirdly, we add a low-pass filter and the edge of the scale in the iterative factor, this way we can get a better result, the image SNR is higher and the MSE is lower. Meanwhile for the image sequence, we adopt the two-step iterative algorithm. The result shows the algorithm own the faster convergence speed and the better convergence result. Different from the other algorithm for blind restoration, although we should select the parameter in the starting of the algorithm, the algorithm doesn’t sensitive for the parameter. So the algorithm possesses very strong adaptability for the blind image deblurring. So a novel algorithm based on an iterative and nonnegative algorithm was proposed to perform blind deconvolution.
Blind Restoration of Motion Blurred Image by Applying a Non-iterative Algorithm
