Optical processing for 3D digital imaging

Conventional optical imaging systems perform both information sensing and image formation functions. The optical system is generally designed to implement processing for image formation with a goal of optimizing analog image quality measures. Digital image involves a fundamental paradigm shift in which the "image" is no longer synonymous with the focal plane field distribution. A digital system may be designed to optimize information transfer from the source to the computational layer without regard to focal plane image quality measures. In this view of the imaging system as a communication channel careful attention must be given to models of how information is encoded in the source space and extracted from the processing system. This paper considers alternative models for source coding and discusses how these models impact information transfer and conventional image quality measures. We also consider system performance when conventional image formation is unnecessary. The shift to digital image dramatically impacts the goals and role of analog processing in optical system components. We describe information theoretic analyses of system performance for coherence and focal imaging of multidimensional spatial and spatio-spectral objects and we present experimental examples for visible and mid-IR radiators.