Computationally Efficient Image Correlation for De-blurring with Photon-Counting Instruments

We present a new implementation of an image alignment and de-blurring algorithm for astronomical observations with photon-counting imagers. The algorithm substantially corrects the effects of observatory attitude changes over long observations and enables the reconstruction of images with superior resolution to those generated without alignment processing. The performance of the algorithm is quantified via simulations of a realistic observation scenario. We formulate a novel, computationally efficient implementation of the alignment algorithm that enables its use for applications with large image sizes and limited computational hardware. Such computational efficiency is critical for spaceflight applications where extreme environments limit the availability of computational hardware to components several generations behind the state of the art. We present an implementation of a Verilog realization of the alignment algorithm which is suitable for execution on a space-qualified Field Programmable Gate Array.