Abstract

Under low light level conditions, photon noise is a fundamental limitation of detected images. Photon noise can be modeled as a Poisson point process given by (1) where f(i, j) is the original image, λ is a proportionality factor determining the SNR, and Poisson(·,·) is a Poisson random number generator. The probabilistic description of a Poisson process is given by (2) The normalized observation is given by (3) Eq. (3) can be rewritten as [1], [2] (4) where (5) is recognized as the noise part of the Poisson process.

© 1986 Optical Society of America