Simulation of extended scenes imaged through turbulence over horizontal paths

We have developed an imaging simulator that accounts for the anisoplanatic effects encountered while imaging extended scenes over horizontal paths. Using an approach that combines geometric and wave optics, an extended scene is divided into discrete point sources. For each point source a ray is traced through discrete Kolmogorovturbulence phase screens toward each pixel in virtual detector. The resulting images express non-uniform tilt and distortion characteristics typical in horizontal surveillance imagery. Using this simulator several large data sets were created based on a known high-resolution source image. By utilizing turbulence corrupted image sets with a known reference image, the performance of image reconstruction techniques can be expressed in terms of common metrics such as Mean Squared Error (MSE). The MSE statistics of a single image corrupted using the simulator over three turbulence conditions are examined relative to a diffraction-limited version of the reference image