Sparse Reconstruction Algorithm Of Total Variation Minimization Based On Alternating Direction Method Of Multipliers

In CT scan,ray has certain harm to the patient,low-dose technique is an important research content in CT technology.low-dose technique can be realized by sparse perspective projection,however,analytsical method will introduce serious artifacts in sparse reconstruction.The reconstruction algorithm based on compression perception can achieve precise sparse reconstruction,and Total Variation minimum is the classic representative of sparse reconstruction.adaptive steepest descent-projection onto convex sets is a kind of effective algorithm that can solve the TV minimum problem,however,because this kind of algorithm has many parameters,we need to choose the best parameters based on experience.Alternating direction method of multipliers can be used to solve the 1 norm problem.In this paper,we research the TV algorithm based on ADMM,the main work is as follows:(1)Designes and implements ADMM based on TV algorithm,compares it with ASD-POCS algorithm and FBP algorithm,and demonstrates TV algorithm ‘s ability of sparse reconstruction(2)Uses the ordered subset algorithm to accelerate ADMM and gives an OS-ADMM algorithm,it can improve the reconstruction speed of ADMM algorithm,in this paper,we focus on the research of the method of the subset partition rule and the projection sequence in the subset(3)Desiends a adaptive method in TV minimum problem and gives an A-ADMM algorithm,this method can obtain high precision reconstruction image.The ADMM algorithm can effectively solve the TV minimization problem,and the algorithm parameters need not be selected manually.In this paper,the comprehensive application of A-ADMM and OS-ADMM algorithm based on the ADMM algorithm will realize the high speed and high precision reconstruction.