Please use this identifier to cite or link to this item:
https://hdl.handle.net/2440/50900
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Tanjung, G. | - |
dc.contributor.author | Lu, T. | - |
dc.contributor.author | Lozo, P. | - |
dc.contributor.author | Liddy, T. | - |
dc.date.issued | 2008 | - |
dc.identifier.citation | Proceedings of the 2008 Australasian Conference on Robotics and Automation / Jonghyuk, K. & Mahony, R. (eds.): pp.1-6 | - |
dc.identifier.isbn | 9780646506432 | - |
dc.identifier.uri | http://hdl.handle.net/2440/50900 | - |
dc.description.abstract | Detection of outdoor wire fences in images is an important and preliminary step in an automated image analysis algorithm for the detection of wire-fence integrity. The challenge is that not only are the wires of the fences in general very thin, a daytime variation in illumination makes the task even more difficult as this may cause non-uniform illumination across the fences. We address the problem of wire-fence detection by an image processing algorithm that combines the Sobel edge operator with an adaptive thresholding technique to generate a binary image. Simulation results demonstrate that the edges of wire fences can be detected accurately under different illumination conditions. | - |
dc.description.statementofresponsibility | Guntur Tanjung, Tien-Fu Lu, Peter Lozo and Tommie Liddy | - |
dc.language.iso | en | - |
dc.publisher | ARAA | - |
dc.source.uri | http://www.araa.asn.au/acra/acra2008/contents.html | - |
dc.title | A robust approach for detecting the edges of outdoor wire fences | - |
dc.type | Conference paper | - |
dc.contributor.conference | Australasian Conference on Robotics and Automation (2008 : Canberra, ACT) | - |
dc.publisher.place | CD | - |
pubs.publication-status | Published | - |
dc.identifier.orcid | Lu, T. [0000-0001-9757-9028] | - |
Appears in Collections: | Aurora harvest 5 Mechanical Engineering conference papers |
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.