Fast and robust deblurring method with multi-frame images based on PSF estimation and total variation optimization

Multi-frame blind restoration algorithms often have disadvantages of slow convergence and sensitiveness to complex noises. In this paper, we proposed a new multi-frame method based on PSF estimate and TV regularization. Our algorithm consists of first simplifying the multi-frame MAP model through the identification of blur parameters for each frame image and then adding various penalty terms to speed up the convergence in deconvolution process obviously. Finally, an adaptive L2/L1 norm selection scheme is built to deal with various noise distributions. A series of experiment results show that the proposed method can outperform other previous work robustly and efficiently even under large noise and image blurring.