'misspelled' visual words in unsupervised range data classification: the effect of noise on classification performance

Recent work in the domain of classification of point clouds has shown that topic models can be suitable tools for inferring class groupings in an unsupervised manner. However, point clouds are frequently subject to non-negligible amounts of sensor noise. In this paper, we analyze the effect on class...

Full description

Saved in:
Bibliographic Details
Published in2011 IEEE/RSJ International Conference on Intelligent Robots and Systems pp. 3850 - 3855
Main Authors Firman, Michael, Julier, Simon
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.09.2011
Subjects
Clustering algorithms
Laser beams
Laser radar
Noise
Solid modeling
Three dimensional displays
Vectors
Online AccessGet full text

Cover

More Information
Summary:Recent work in the domain of classification of point clouds has shown that topic models can be suitable tools for inferring class groupings in an unsupervised manner. However, point clouds are frequently subject to non-negligible amounts of sensor noise. In this paper, we analyze the effect on classification accuracy of noise added to both an artificial data set and data collected from a Light Detection and Ranging (LiDAR) scanner, and show that topic models are less robust to 'misspelled' words than the more näive k-means classifier. Furthermore, standard spin images prove to be a more robust feature under noise than their derivative, 'angular' spin images. We additionally show that only a small subset of local features are required in order to give comparable classification accuracy to a full feature set.
ISBN:1612844545
9781612844541
ISSN:2153-0858
DOI:10.1109/IROS.2011.6095016