Iterative Image Deblurring Algorithm using Complementary Pair of Filters

In this paper, an iterative algorithm for image deblurring is proposed. The algorithm is based on the complementary pair of filters. The modified Tikhonov regularization is used to control the optimized iteration of the deblurring process. Several sizes of Gaussian blur are applied to produce the degraded images in our experiments. The Peak Signal to Noise Ratio (PSNR) is used to measure the image deblurring quality. The results show that the proposed algorithm is superior to the traditional Lucy-Richardson algorithm and regularized filter both visualized observation and a quality metric. It performs well on any size of the blur kernel and can be applied in the non-exact blur kernel. Moreover, the edge ringing and boundary artifact are unnoticeable.