Fundamental Performance Bounds of Phase Diversity Blind Deconvolution Algorithm for Various Diversity Polynomials, Noise Statistics, and Scene Size

摘要

In this paper, we will present information theoretic bounds on the estimated Zernike coefficients for various diversity phase functions. We will show that, in certain cases, defocus diversity may yield higher Cramer-Rao lower bound (CRLB) than some other diversity phase functions. Evaluating the performance of the phase diversity algorithm using simulated images, we find that for an extended scene and defocus diversity, the phase diversity algorithm achieves the CRLB for known objects and approaches the CRLB by about a factor of two for unknown objects.