Artificial Neural Nets Object Recognition for 3D Point Clouds

Author

Habermann, Danilo ; Hata, Alberto ; Wolf, Denis ; Osorio, Fernando Santos

Author_Institution

Mobile Robot. Lab.-LRM, Univ. of Sao Paulo, São Carlos, Brazil

fYear

2013

fDate

19-24 Oct. 2013

Firstpage

101

Lastpage

106

Abstract

This paper presents a approach that uses 3D point clouds from laser sensor to perform the classification of typical obstacles in urban environments. The presented method consists of Velodyne Lidar point clouds segmentation, feature extraction and use of a MLP Neural Nets to classify vehicles, people, tree trunks, light poles and buildings. Experimental results demonstrated that is possible to recognize different classes of 3D structures with a very good precision. At the end, the performances of two neural networks are compared.

Keywords

feature extraction; image classification; image segmentation; image sensors; multilayer perceptrons; object recognition; optical radar; solid modelling; 3D point clouds; 3D structures recognition; MLP neural nets; Velodyne Lidar point clouds segmentation; artificial neural nets object recognition; buildings classification; feature extraction; laser sensor; light poles classification; obstacles classification; people classification; tree trunks classification; urban environments; vehicles classification; Buildings; Image segmentation; Navigation; Neural networks; Robot sensing systems; Three-dimensional displays; Vehicles; lidar point clouds; neural network; object classification; segmentation;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Systems (BRACIS), 2013 Brazilian Conference on

Conference_Location

Fortaleza

Type

conf

DOI

10.1109/BRACIS.2013.25

Filename

6726433