AN ITERATIVE METHOD TO SOLVING THE PROBLEM OF RECONSTRUCTING A 2-TENSOR FIELD FROM LIMITED DATA

In this paper, we consider problem of recovery of a two-dimensional symmetric 2-tensor field. Data are values of ray transform in domain limited by angles. We apply the iterative Gerchberg-Papoulis algorithm to this problem. Steps of the algorithm are performed in two spaces: the main space of tensor field components and the Fourier space. The actions performed are based on back-projection, direct and inverse Fourier transforms and the use of a priori information. Numerical experiments on recovery of test 2-tensor fields showed good results.