Tentative Tests on Two Rapid Multispectral Classifiers for Classifying Point Clouds

Tentative Tests on Two Rapid Multispectral Classifiers for Classifying Point Clouds

Zheng, M. (TU Delft OLD Department of GIS Technology; Wuhan University)
Lemmens, M.J.P.M. (TU Delft OLD Department of GIS Technology)
van Oosterom, P.J.M. (TU Delft OLD Department of GIS Technology)

Bregt, Arnold (editor)
Sarjakoski, Tapani (editor)
Lammeren, Ron van (editor)
Rip, Frans (editor)

2017

This paper focusses on the feasibility of classifiers, developed for classifying multispectral images, for assigning classes to point clouds of urban scenes. The motivation of our research is that dense point clouds require fast classification methods to extract meaningful information within a reasonable amount of time and multispectral classifiers do have this property. We employ two encoding methods acting on one feature: the altitude above street level. We emphasize computation time and therefore we use just one feature in this prelimina

Classification
point clouds
feature encoding

http://resolver.tudelft.nl/uuid:910008bd-8904-4356-b5f3-49f2b33f3b09

Wageningen University

978-90-816960-7-4

Proceedings of the 20th AGILE Conference on Geographic Information Science: Societal Geo-innovation

AGILE 2017: 20th AGILE International Conference on Geographic Information Science, 2017-05-09 → 2017-05-12, Wageningen, Netherlands

Institutional Repository

conference paper

© 2017 M. Zheng, M.J.P.M. Lemmens, P.J.M. van Oosterom