IMT Atlantique - IMT Atlantique Bretagne-Pays de la Loire (Campus Brest : Technopôle Brest-Iroise CS 8381829238 BREST Cedex 3 -
Campus Nantes : 4, rue Alfred Kastler- La chantrerie 44300 NANTES -
Campus Rennes : 2 Rue de la Châtaigneraie, 35510 CESSON SEVIGNE - France)
Abstract : This paper introduces a new definition of multiscale neighborhoods in 3D point clouds. This definition, based on spherical neighborhoods and proportional subsampling, allows the computation of features with a consistent geometrical meaning, which is not the case when using k-nearest neighbors. With an appropriate learning strategy, the proposed features can be used in a random forest to classify 3D points. In this semantic classification task, we show that our multiscale features outperform state-of-the-art features using the same experimental conditions. Furthermore, their classification power competes with more elaborate classification approaches including Deep Learning methods.
Type de document :
Communication dans un congrès
3DV 2018 - International Conference on 3D Vision, Sep 2018, Verone, Italy
https://hal.archives-ouvertes.fr/hal-01852565
Contributeur : Hugues Thomas
<>
Soumis le : mercredi 1 août 2018 - 20:34:05
Dernière modification le : lundi 12 novembre 2018 - 11:03:10
Document(s) archivé(s) le : vendredi 2 novembre 2018 - 14:33:04
Hugues Thomas, Jean-Emmanuel Deschaud, Beatriz Marcotegui, Francois Goulette, Yann Le Gall. Semantic Classification of 3D Point Clouds with Multiscale Spherical Neighborhoods. 3DV 2018 - International Conference on 3D Vision, Sep 2018, Verone, Italy. 〈hal-01852565〉
