Superresolution with an Opacity Constraint

Abstract

The term superresolution can be defined as the use of a priori information to achieve resolution superior to the diffraction-limited resolution. In this paper we explore the use of a novel type of prior knowledge in the context of 3-D imaging. We wish to exploit knowledge that the object being imaged is confined to a 2-D manifold (surface) embedded in a 3-D space. Such prior knowledge is valid when the object is opaque to the illuminating radiation so that only secondary sources that lie in the outer surface of the object contribute to the reflected field.