Fast computation of readout smear correction for framestore CCD based images

Shutterless operation of cameras with framestore Charge-Coupled-Devices (CCDs) produces readout smear as characteristic image degradation. Post-facto image corrections for readout smear are known for the standard case, the case of charge flush and reverse clocking operations, the periodic case, and the aperiodic case. This paper gives an overview of these state of the art algorithms. It is shown, that the correction in the standard and in the periodic case are convolution operations, which allow for faster computations and offer transfer functions. Algorithms for parallel computation of readout smear corrections are derived from the case for charge flush and reverse clocking. The aperiodic case cannot be calculated via convolutions and is approximated by the periodic case. We finally introduce a parallel and iterative variant of this approximation, which allows for the use of unshuttered framestore CCDs on autonomous astronomical observations.