Optical EngineeringParameterized blind deconvolution of laser radar imagery using an anisoplanatic optical transfer function
|Format||Member Price||Non-Member Price|
Gated laser radar imaging systems hold unique promise for long-distance remote sensing applications. Short-exposure speckle imagery from wide field-of-view (FOV) systems may be used to jointly estimate maximum likelihood estimates of the remote scene together with the atmospheric seeing condition parameterized by Fried's parameter. Previous research has indicated that employment of the short-exposure optical transfer function (OTF) within the deconvolution algorithm yields slightly pessimistic estimates of Fried's parameter. It was postulated that the short-exposure OTF included excessive high spatial frequency components when applied to wide FOV systems, yielding seeing condition estimates of the imaging scenario that were slightly lower than if the atmospheric conditions admitted an anisoplanatic system imaging model. To better estimate Fried's parameter, an anisoplanatic OTF (AOTF) was developed using a tilt-only phase correlation approximation. This AOTF was used together with the short-exposure OTF within the MAP algorithm, and estimated seeing conditions were compared for both simulated and experimental wide FOV scenarios. It was found that the additional anisoplanatic blur components modeled by the AOTF increased the accuracy of the estimation of Fried's parameter from 5% to within 2% using simulated imagery, and from 8.6% to within 2.9% using experimentally collected image data.