Based on a standard binomial model of sub-Poissonian photocounting statistics, we analyze the discrimination performance between the two possibilities that light has been potentially absorbed or not. For that purpose, we study with numerical simulations the behavior of different information-theory-based measures of the contrast and show that the Chernoff measure allows one to obtain a useful contrast characterization that has simple physical interpretation and that helps in analyzing the benefit of using sub-Poissonian light to improve detection tasks.
© 2009 Optical Society of America
Original Manuscript: December 10, 2008
Manuscript Accepted: February 20, 2009
Published: April 8, 2009
Julien Fade, Philippe Réfrégier, Nicolas Treps, and Claude Fabre, "A gain criterion for the improvement of detection tasks with sub-Poissonian light," J. Opt. Soc. Am. A 26, 1139-1146 (2009)