Super-Resolution restoration technique for ground-based imaging of space objects

Space objects images taken through large ground-based telescopes usually suffer from a degradation due to atmospheric turbulence, In order to reduce the cost of Charge-coupled Device (CCD) and improve the image signal-to-noise ratio, ground-based telescopes are usually designed with down-sampling, the observed image is blurry and aliased. We present a Super-Resolution (SR) algorithm to restore under-sampled image sequences with randomly varying blur, the algorithm significantly improves the quality and resolution of space object images degraded by atmospheric turbulence, it is a unifying framework that simultaneously performs Multi-frame Blind Deconvolution (MFBD) and SR in a maximum a posteriori (MAP) framework. The object and the Point Spread Function (PSF) are estimated by minimizing a cost function coming from the MAP criteria, the Total Variation (TV) regularization is imposed on the object estimation, TV regularization is remarkably effective to suppress the noise and to preserve the sharp edges in the image. We use the conjugate gradient method for the minimization for its fast convergence. Encouraging simulation results demonstrate that the restored image produced by this algorithm often have better quality than MFBD.