Andreas Winkler, Bernhard Zagar,
"Robust Estimation of Geometric Parameters from a Single Image in Support of Steel Coil Quality Assurance"
, in Rudolf Scheidl and Bernhard Jakoby: The 13th Mechatronics Forum International Conference Proceedings, Vol. 1, Trauner, Linz, Seite(n) 55-62, 9-2012, ISBN: 978-3-99033-046-1
Robust Estimation of Geometric Parameters from a Single Image in Support of Steel Coil Quality Assurance
Sprache des Titels:
The 13th Mechatronics Forum International Conference Proceedings
Although steel coils are big, heavy and clumsy products, the customers put very high demands on their quality. The surface quality is of interest particularly, as it is considered as an indicator for the state of the whole coil. When it comes down to detecting surface defects like scratches, dents or packaging errors, optical inspection by a digital camera seems to be the first choice. However, acquiring the whole surface of a cylindrical coil turns out to be a difficult task, especially in an industrial environment and in view of the fact that coils typically weigh about 30 tons. To get an image of the entire cylinder barrel out of a few acquisitions around the coil, it is necessary to develop a method to measure the geometric parameters of the coil and the realtive position of the camera. The proposed procedure attempts to estimate these quantities out of single images, using detected contours and some prior knowledge about the geometry (the coil is assumed to be an ideal cylinder). It is intended to generalize this method even to other solids, which can not be characterized simply by identification of straight lines.