Deautoconvolution in the two-dimensional case (Q6163341)

This paper studies the deautoconvolution in the two-dimensional case both analytically and numerically. The problem is about the reconstruction of a real function of two real variables over the unit square from observations of its autoconvolution on \([0; 2]^2\) (full data case) or on \([0; 1]^2\) (limited data case). The authors discuss the uniqueness and non-uniqueness for the deautoconvolution problem, and characterize its ill-posedness. Extensive numerical illustrations are given for Tikhonov-type regularization with different penalties and iteratively regularized Gauss-Newton method.