Abstract

Many remote sensing systems are undersampled, which traditionally precluded their use with phase diversity algorithms. Phase-diverse phase retrieval (PDPR) algorithms, which assume a point object, have been generalized to deal with the undersampled case by including a number of undersampled, spatially-displaced point source images within the nonlinear optimization. A different approach is presented in which super-resolution is used to generate Nyquist-sampled images prior to phase diversity reconstruction. Experimental results are presented for two PDPR algorithms, but the technique is also extensible to phase diversity imaging.

© 2012 Optical Society of America