Neural networks for multi-scale edge classification in 3D point cloud
1 : Institute de recherche en informatique de Toulouse
(IRIT)
Université Toulouse III - Paul Sabatier
Point clouds are unstructured clusters of data representing geometries. Point cloud analysis is a challenging field, especially challenging for use of neural networks. In this presentation I will explore point cloud analysis using neural networks, and the combination between scale-space analysis and neural network classifier, to make a robust and fast edge detection network for point cloud labeling.