Abstract
Estimation of surface curvature from range data is important for a range of tasks in computer vision and robotics, object segmentation, object recognition and robotic grasping estimation. This work presents a fast method of robustly computing accurate metric principal curvature values from noisy point clouds which was implemented on GPU. In contrast to existing readily available solutions which first differentiate the surface to estimate surface normals and then differentiate these to obtain curvature, amplifying noise, our method iteratively fits parabolic quadric surface patches to the data. Additionally previous methods with a similar formulation use less robust techniques less applicable to a high noise sensor. We demonstrate that our method is fast and provides better curvature estimates than existing techniques. In particular we compare our method to several alternatives to demonstrate the improvement.
| Original language | English |
|---|---|
| Title of host publication | Australasian Conference on Robotics and Automation, ACRA 2015 |
| Subtitle of host publication | 2-4 December 2015; Canberra, Australia |
| Editors | Robert Mahony, Jonghyuk Kim, Hongdong Li |
| Place of Publication | Canberra ACT Australia |
| Publisher | Australian Robotics & Automation Association Inc |
| Pages | 33-41 |
| Number of pages | 9 |
| ISBN (Electronic) | 9780980740462 |
| ISBN (Print) | 9781510819269 |
| Publication status | Published - 2015 |
| Event | Australasian Conference on Robotics and Automation 2015 - Australian National University (ANU), Canberra, Australia Duration: 2 Dec 2015 → 4 Dec 2015 http://www.araa.asn.au/conferences/acra-2015/ |
Conference
| Conference | Australasian Conference on Robotics and Automation 2015 |
|---|---|
| Abbreviated title | ACRA 2015 |
| Country | Australia |
| City | Canberra |
| Period | 2/12/15 → 4/12/15 |
| Internet address |
Cite this
}
A fast method for computing principal curvatures from range images. / Spek, Andrew Tristan; Li, Wai Ho; Drummond, Thomas William.
Australasian Conference on Robotics and Automation, ACRA 2015: 2-4 December 2015; Canberra, Australia. ed. / Robert Mahony; Jonghyuk Kim; Hongdong Li. Canberra ACT Australia : Australian Robotics & Automation Association Inc, 2015. p. 33-41.Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research › peer-review
TY - GEN
T1 - A fast method for computing principal curvatures from range images
AU - Spek, Andrew Tristan
AU - Li, Wai Ho
AU - Drummond, Thomas William
PY - 2015
Y1 - 2015
N2 - Estimation of surface curvature from range data is important for a range of tasks in computer vision and robotics, object segmentation, object recognition and robotic grasping estimation. This work presents a fast method of robustly computing accurate metric principal curvature values from noisy point clouds which was implemented on GPU. In contrast to existing readily available solutions which first differentiate the surface to estimate surface normals and then differentiate these to obtain curvature, amplifying noise, our method iteratively fits parabolic quadric surface patches to the data. Additionally previous methods with a similar formulation use less robust techniques less applicable to a high noise sensor. We demonstrate that our method is fast and provides better curvature estimates than existing techniques. In particular we compare our method to several alternatives to demonstrate the improvement.
AB - Estimation of surface curvature from range data is important for a range of tasks in computer vision and robotics, object segmentation, object recognition and robotic grasping estimation. This work presents a fast method of robustly computing accurate metric principal curvature values from noisy point clouds which was implemented on GPU. In contrast to existing readily available solutions which first differentiate the surface to estimate surface normals and then differentiate these to obtain curvature, amplifying noise, our method iteratively fits parabolic quadric surface patches to the data. Additionally previous methods with a similar formulation use less robust techniques less applicable to a high noise sensor. We demonstrate that our method is fast and provides better curvature estimates than existing techniques. In particular we compare our method to several alternatives to demonstrate the improvement.
UR - http://www.scopus.com/inward/record.url?scp=85018494191&partnerID=8YFLogxK
UR - http://www.araa.asn.au/acra/acra2015/papers/pap117.pdf
M3 - Conference Paper
SN - 9781510819269
SP - 33
EP - 41
BT - Australasian Conference on Robotics and Automation, ACRA 2015
A2 - Mahony, Robert
A2 - Kim, Jonghyuk
A2 - Li, Hongdong
PB - Australian Robotics & Automation Association Inc
CY - Canberra ACT Australia
ER -