An Innovative Image Reconstruction Method for High-Precision Astrometry

An automatic image-reconstruction method is presented, based on the use of a mathematical morphology operator. Variable structural elements along multiple directions are adopted for image convolution, and then all of the corresponding convolved images are stacked to obtain a final result. To investigate its efficiency, both simulated images and raw CCD images, produced by a trial observation using a 45-cm telescope, are tested. With this method, the influence of degradation in astrometric CCD images is reduced, and the position accuracy of background stars is distinctly improved. This method will contribute significantly to optical data reduction and high-order precision astrometry.