Date of Award


Document Type


Degree Name

Master of Science


Department of Engineering Physics

First Advisor

Michael C. Roggemann, PhD


The U.S. Air Force uses adaptive optics systems to collect images of extended objects beyond the atmosphere. These systems use wavefront sensors and deformable mirrors to compensate for atmospheric turbulence induced aberrations. Adaptive optics greatly enhance image quality; however, wavefront aberrations are not completely eliminated. Therefore, post-detection processing techniques are employed to further improve the compensated images. Typically, many short exposure images are collected, recentered to compensate for tilt, and then averaged to overcome randomness in the images and improve signal-to-noise ratio. Experience shows that some short exposure images in a data set are better than others. Frame selection exploits this fact by using a quality metric to discard low quality frames. A composite image is then created by averaging only the best frames. Performance limits associated with the frame selection technique are investigated in this thesis. Limits imposed by photon noise result in a minimum object brightness of visual magnitude +8 for point sources and +4 for a typical satellite model. Effective average point spread functions for point source and extended objects after frame selection processing are almost identical across a wide range of conditions. This discovery allows the use of deconvolution techniques to sharpen images after using the frame selection technique. A new post-detection processing method, frame weighting, is investigated and may offer some improvement for dim objects during poor atmospheric seeing. Frame selection is demonstrated for the first time on actual imagery from an adaptive optics system. Data analysis indicates that signal-to-noise ratio improvements are degraded for exposure times longer than that allowed to "freeze" individual realizations of the turbulence effects.

AFIT Designator


DTIC Accession Number



The author's Vita page is omitted.