Artificial neural networks for noisy image super-resolution

Noisy incoherent objects, which are too close to be remotely separated by o ptically imaging beyond the Rayleigh diffraction limit, might be resolved b y employing the artificial neural network (ANN) smart pixel post-processing and its mathematical framework, independent component analysis (ICA). It i s shown that ICA ANN approach to superresolution based on information maxim ization principle could be seen as a part of the general approach called sp ace-bandwidth product adaptation method. Our success is perhaps due to the blind source separation smart-pixel detectors behind the imaging lens (inve rse adaptation), while the Rayleigh diffraction limit remains valid for a s ingle instance of the deterministic imaging systems' realization. The blind ness is due to the unknown objects, and the unpredictable propagation effec t on the net imaging point spread function, Such a software/firmware enhanc ement of imaging system may have a profound implication to the designs of t he new (third) generation imaging systems as well as other nonoptical imagi ng systems. (C) 2001 Elsevier Science B.V. All rights reserved.