Automatic Detection and Classification of Defect on road Pavement using Anisotropy Measure
Résumé
Automatic pavement cracking detection is a part of road maintenance and rehabilitation strategies. Cracks detec-tion is one of the main features used by road authorities to manage efficiently its networks. Road surface is made using aggregates which can have different sizes, organized randomly. Scanned pictures of theses surfaces appear has random distribution of a re-duced set of gray levels. Cracks or defaults can't be ex-tracted by a simple threshold. In this paper, we introduce a measure of anisotropy for characterization of cracks. Fundamental idea of this meth-od is to detect the variation of features by considering different orientations. It supposes that a defect-free texture is quasi-homogeneous according to each orientation, while a cracked texture presents differences to, at least, one direction. Comparative results of anisotropy method with threshold method and 2D wavelet transform method are presented to illustrates benefits of anisotropy. We show that the method can be used to detect others types of defects, such as joints or on other surfaces such as ceramic, granite slabs, that may offer perspectives for quality control in industry.
Domaines
Automatique / Robotique
Origine : Fichiers produits par l'(les) auteur(s)
Loading...