Joint image and pupil plane reconstruction algorithm based on Bayesian techniques

The focus of this research was to develop an joint pupil and focal plane image recovery algorithm for use with coherent LADAR systems. The benefits of such a system would include increased resolution with little or no increase in system weight and volume as well as allowing for operation in the absence of natural light since the target of interest would be actively illuminated. Since a pupil plane collection aperture can be conformal, such a system would also potentially allow for the formation of large synthetic apertures.; The algorithm developed used many frames of coherent pupil and focal plane data. The data frames are summed in the respective planes to give two data set (one for each plane). Appropriate statistical models are used and a joint Maximum Likelihood estimator is formed. The algorithm is tested using a Monte Carlo approach. The system is demonstrated to be robust and in all but extreme cases yield better results than algorithms using a single data set (such as deconvolution). It was shown that the joint algorithm had a resolution increase of 70% over deconvolution alone and a 40% increase over traditional pupil plane algorithms. It is also demonstrated that the new algorithm does not suffer as severely from stagnation problems typical with pupil plane algorithms. A stopping criteria based on the statistics of the data was also developed.