| Title |
Design of 3D scene scanner for flat surface detection / |
| Authors |
Lipnickas, A ; Rimkus, K ; Sinkevičius, S |
| DOI |
10.1007/978-3-319-06883-1_2 |
| ISBN |
9783319068824 |
| eISBN |
9783319068831 |
| Full Text |
|
| Is Part of |
Issues and challenges in artificial intelligence / Z.S. Hippe, J.L. Kulikowski, T. Mroczek, J. Wtorek (eds).. Cham : Springer, 2014. p. 17-26.. ISBN 9783319068824. eISBN 9783319068831 |
| Abstract [eng] |
This chapter describes the design of a 3D space scene scanning system built from a 2D laser scanner merged with a CCD colour camera; it also presents an algorithm for flat area detection in a 3D point cloud. For that purpose, the RANdom SAmple Consensus (RANSAC) search engine has been adopted for flat area segmentation and planes detection. Due to a fact that human made planes are limited in size, we have proposed data filtering by comparing averaged point triangulation normals to the selected plane normal. The experiments have shown good results for an analysed environment segmentation with the applied angle variation measure up to ±25°. The applied variation threshold allowed to segment flat planes areas considering surface curvedness. |
| Published |
Cham : Springer, 2014 |
| Type |
Book part |
| Language |
English |
| Publication date |
2014 |
