Duan, J;
Lachhani, K;
Baghsiahi, H;
Willman, E;
Selviah, DR;
(2014)
Indoor rigid sphere recognition based on 3D point cloud data.
In: Coleman, S and Gardiner, B and Kerr, D, (eds.)
Proceedings: IMVIP 2014: 2014 Irish Machine Vision and Image Processing.
(pp. pp. 28-33).
Irish Pattern Recognition & Classification Society: Derry-Londonderry, UK.
![[thumbnail of Indoor rigid sphere recognition based on 3D point cloud data （final)-1.pdf]](https://discovery.ucl.ac.uk/style/images/fileicons/text.png) |
Text
Indoor rigid sphere recognition based on 3D point cloud data （final)-1.pdf Download (209kB) |
Abstract
This paper presents a method for recognising spherical shapes in 3D point cloud XYZ coordinate data obtained by scanning an indoor environment using a LIDAR scanner. Firstly, bilateral smoothing is performed to smooth the surfaces consisting of points. Then, the surface curvature and surface roughness of each point in the scan are extracted by analysing the point cloud data. Finally, a three layer multilayer perceptron neural network trained by the Levenberg-Marquardt algorithm is used to automatically distinguish points belonging to spheres from all the other points making use of extracted features. A novel feedback technique is applied in which the neural network is used several times on the recognised data. This method can be applied to automate 3D scan alignment.
| Type: | Proceedings paper |
|---|---|
| Title: | Indoor rigid sphere recognition based on 3D point cloud data |
| Event: | Irish Machine Vision and Image Processing Conference |
| Location: | Derry-Londonderry, Northern Ireland |
| Dates: | 2014-08-27 - 2014-08-29 |
| Open access status: | An open access version is available from UCL Discovery |
| Publisher version: | http://hdl.handle.net/2262/71411 |
| Language: | English |
| Additional information: | Published by the Irish Pattern Recognition and Classification Society (IPRCS), a member body of the International Association for Pattern Recognition (IAPR). This paper is published under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Licence (http://creativecommons.org/licenses/by-nc-sa/4.0/). |
| Keywords: | Point Cloud Data, Object Recognition, Neural Networks, 2D Laser Scanning, LIDAR |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Electronic and Electrical Eng UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Maths and Physical Sciences |
| URI: | https://discovery.ucl.ac.uk/id/eprint/1434810 |
Archive Staff Only
 |
View Item |
