A New Method for NMR Inversion Constrained by Generalized Prior Information

A new method for NMR inversion constrained by generalized prior information was proposed. Power function transform and exponential power function transform were used to directly extract generalized prior information from echo data. Then the prior information was used as constraint conditions to construct the new objective function and from which we got the inversion results. The numerical simulation results showed that the new method improved the inversion accuracy for T2 inversion and T2-D inversion at low SNR level compared with traditional regularization method.