On approximating mathematical morphology operators via deep learning techniques
Résumé
Mathematical Morphology (MM) is a well-established discipline whose aim is mainly to provide tools to characterise complex object via their shape/size features. This study addresses the problem of robust approximation of mathematical morphology (MM) operators by deep learning methods. We present two cases, (a) Asymmetric autoencoders for part-based approximations of classical MM in the sense of [1] and, (b) image-to-image translation networks [2] to produce robust MM operators in presence of noise.