Volumetric analysis of digital objects using distance transformation: performance issues and applications
Résumé
In binary images, the distance transformation (DT) and the geometrical medial
axis are classic tools for shape analysis. In the digital geometry literature, recent articles
have demonstrated that fast algorithms can be designed without any approximation of the
Euclidean metric. The aim of the paper is to first give an overview of separable techniques
to compute the distance transformation, the reverse distance transformation and a discrete
medial axis extraction with the Euclidean metric. Then we will focus on performance issues
and different applications in shape analysis.