CLEANING ASTRONOMICAL DATABASES USING HOUGH TRANSFORMS AND RENEWAL STRINGS

摘要

In this paper we are concerned with detecting artefactual linear features such as satellite or aeroplane tracks and scratches in astronomical images,. The standard approach to line detection is the Hough transform (HT) By combining the HT with renewal processes and hidden Markov models we obtain a probabilistic model (renewal strings) that'is a highly effective method for removing these artefacts.