The Bregman Methed Research on Image Reconstruction Based on Compressive Sensing

杨晓兰,朱永贵,丛佳,刘平,李雪萍

DOI: https://doi.org/10.3969/j.issn.1673-4793.2012.04.002

2012-01-01

Abstract:Because information such as boundaries of organs is very sparse in most MR images,compressed sensing makes it possible to reconstruct the same MR images from a very limited set of measurements significantly reducing the MRI scan duration.In order to do that,however,one has to solve the difficult problem of minimizing nonsmooth functions on large data sets.To handle this,we propose an efficient algorithm that It has overcome the computational complexity of solving the l1 problem,this paper puts forward β norm approximation l1 norm thought,because β norm with slickness,one can use Bregman iterative regularization method to solve this problem.The numerical experiments demonstrate that original MR images can be reconstructed exactly from the mere 40 percent of the complete set of measurements by our approach.