Abstract

A powerful mathematical and computational technique for the solution of a number of experimental inverse problems, described by an underlying first-kind Fredholm equation has been under development by several groups in recent years. The programme had its origins in original theoretical work by McWhirter and Pike [1] in which, for the first time, the inverse Laplace transform was solved by ‘information theory’ techniques. These techniques have been greatly generalised by the use of the theory of singular systems which provide exact ‘Fourier’ orthonormal bases for the description of experimental data and reconstructed continuous object solutions of the data processing problem.

© 1992 Optical Society of America