Sparse projection CT image reconstruction based on the split Bregman less iteration

Abstract

Sparse angle projection CT image reconstruction in medical diagnosis and industrial non-destructive testing has important theoretical significance and practical application value. In the paper, L1 norm was introduced as the CT images of regular constraint and optimization reconstruction model, and the method to solve it was presented based on the Split Bregman algorithm. Shepp-Logan numerical simulation experiments show that the image reconstructed by the traditional algebraic reconstruction algorithm of ART for sparse projection CT is poor. The Split Bregman may solve L1 regularization constraint model of sparse projection of CT with less number of iterations, fast reconstruction and good reconstruction quality. For the splitting factor of the algorithm, in a numerical range, the greater the reconstruction quality is better.