Research On Reweighting-based Edge-Preserving Algorithm For Limited-Angle Image Reconstruction

Computed Tomography（CT）has many advantages such as non-destructive,high-resolution,and no overlapping images,and is widely used in security,medical and industrial fields.Generally,traditional CT systems require full-angle projection data for image reconstruction.However,due to the limitation of practical conditions,sometimes it is impossible to obtain full-angle projection data,which leads to the problem of CT image reconstruction with limited-angle projection data.Theoretically,the problem of limited-angle image reconstruction is an ill-posed problem,and this ill-posedness will become more serious as the missing angles increasing.Therefore,it is known that the reconstruction of limited-angle images with small angles is a very challenging problem.Regarding of the problem of limited-angle image reconstruction with known small angles,the severe ill-posedness of the theory makes the corresponding discrete linear equations=(7 also severely ill-posed.This severe ill-posedness is caused by the very large condition number of.We introduce a weighting method to improve this condition number,thereby improving the well-posedness,and combining the weighting method with the existing edge-preserving continuation and edge-preserving smoothing algorithms,we propose a limited-angle reconstruction model,which is solved by an alternating method.Numerical experiments show that the algorithm proposed in this paper has better reconstruction results than existing algorithms at the same limited angle.