A Regularized Reconstruction Method for Optical Tomography Based on Conjugate-gradient

Jiajun Wang
2006-01-01
Abstract:A new regularized conjugate-gradient reconstruction algorithm for optical tomograpy based on diffusion equation is presented in this article. In order to improve the ill-poseness in reconstruction process, two criterions with smoothness -the image entropy and the local smoothing function are incorporated as the regularized items. Compared with the reconstruction without regularized items, the MSE of the image are 0.4413×10-4 and 0.4819×10-4. The results show that this regularized reconstruction algorithm can effectively decrease the ill-posedness of the reconstruction and improve reconstruction quality.
What problem does this paper attempt to address?