Reconstruction of 3D reflective index field from limited orthogonal projections using expansion of local basic function

Image reconstruction by sparse data is an important and intractable problem in the field of OCT (optical computerized tomography). On the basis of previous work on the measurement of refractive index distribution by phase-shifting interferometry, the feasibility of reconstruction from orthogonal projections within limited narrow views is discussed. Local differentiable basic function was used to expand the refractive index filed, and consequently for precision in reconstruction, an iterative reconstruction algorithm with multiple-changeable relaxation factors is presented, based on genetic algorithm. Experiments and numerical simulation showed that the method of reconstruction from a few orthogonal projections within limited views caused a marked reduction in the quantity of measurement data. It is proved that the application of multi-changeable relaxation factors to the process of iteration makes the reconstruction precision considerably improved reaching a max error of no more than 5% and an average error less than 1%, when compared with the conventional grid method, associated with the expansion of local differentiable basic function.