Limited-angle CT reconstruction with primal-dual algorithm
作者单位：School of Mathematical SciencesCapital Normal University
会议名称：《第十七届中国体视学与图像分析学术会议》
会议日期：2022年
学科分类：08[工学] 080203[工学-机械设计及理论] 0802[工学-机械工程]
关 键 词：onvex and nonconvex regularization primal-dual,limited-angle CT
摘 要：Due to the consideration of X-ray dose and physical restrictions of scanning environment or the object,the acquired projection data usually are not *** algorithms like ART,FBP with limited-angle data will introduce blur and streak *** improve the reconstruction quality,many models and algorithms that incorporate various regularizers,like total variation(TV),i.e.l norm,l norm and l norm of gradient,have been ***,the directional TV model has been proposed and demonstrated superior reconstruction abilities with limited-angle *** motivates us to study the capacity of convex models with primal-dual *** this work,a study on convex and nonconvex models solved by the Chambolle-Pock primal-dual algorithm for limited-angle CT problem is carried *** investigation includes a set of optimization models including constrained TV and its reformulations as well as nonconvex l norm regularizers.