Abstract

A typical wave front sensor in an adaptive optical imaging system requires a point source or a glint on the object to operate, as do the various image-based phase retrieval methods. However, with the image sharpening method,^1,2 one can sense the aberrations from the aberrated image of a point or extended object. This is done by introducing aberrations in various modes with an adaptive mirror to maximize the image sharpness. We describe simulations of image sharpening in single and multiaperture systems imaging extended objects. In simulating a multiaperture system the algorithm has been successfully applied to phase the elements of an array by correcting their piston and tilt errors. We define a noise equivalent aberration as the amount required to discern changes in the sharpness function over the system noise and describe a relationship between it and the image signal-to-noise ratio. We determine the efficiency of various optimization schemes to maximize the sharpness such as successive linear searches and multidimensional gradient searches. The algorithm performance with different amounts of photon and additive noise is also evaluated.

© 1985 Optical Society of America