An Adaptive Algorithm Based On L_q(q=1/2,1,2) Regularization For Image Restoration

In this paper,we present an adaptive algorithm based on L_q(q=1/2,1,2)regu-larization for image restoration.We adopt splitting Bregman method and alternating minimization methods to solve the generalized ROF model.The main idea of splitting Bregman method is to transform complicated nonconvex minimization problem into two simple subproblems.To improve quality of the image and the speed of computing,we adaptively select the values of q(q=1/2,1,2) according to the gradient informa-tion of each pixel image.The restoration image can not only keep the original edge and original detail information but also weaken the stair phenomenon.The numerical experiments demonstrate that our adaptive algorithm is efficient and robust even for images with larger noises and blur operators.