Improved laser scan for pitting corrosion measurement by using super resolution technique

In this paper, a single-frame super resolution method is proposed to infer a high-resolution laser scan from the low-resolution input. The relation between the low resolution input and high-resolution result is modeled with a Markov random field (MRF) with the aid of a training set built in advance. A belief propagation algorithm is implemented to infer the super-resolved result. The experiments demonstrate a good performance of the proposed method in comparison with the traditional interpolation methods.Dans ce document, une m\ue9thode de super-r\ue9solution image par image est propos\ue9e pour inf\ue9rer un balayage laser haute r\ue9solution \ue0 partir d?une entr\ue9e basse r\ue9solution. La relation entre l?entr\ue9e basse r\ue9solution et le r\ue9sultat haute r\ue9solution est mod\ue9lis\ue9e avec un champ MRF, \ue0 l?aide d?une base d?apprentissage construite au pr\ue9alable. Un algorithme de propagation des croyances est mis en ?uvre pou inf\ue9rer le r\ue9sultat super-r\ue9solu. Ces exp\ue9riences d\ue9montrent la performance satisfaisante de la m\ue9thode propos\ue9e par rapport aux m\ue9thodes d?interpolation classiques.Peer reviewed: YesNRC publication: Ye