Title: Strip wrinkling detection based on feature extraction and sparse representation

Authors: Wenhao Wang; Xiong Chen; Yuqi Pan

Addresses: Intelligent Control Research Lab, Department of Electronic Engineering, Fudan University, Shanghai 200433, China ' Intelligent Control Research Lab, Department of Electronic Engineering, Fudan University, Shanghai 200433, China ' Intelligent Control Research Lab, Department of Electronic Engineering, Fudan University, Shanghai 200433, China

Abstract: Strip wrinkling is a kind of adverse phenomenon occurring during strip steel production. It is necessary to detect strip wrinkling promptly for production lines, most of which rely on manual survey for detection until now. To provide a better condition for wrinkling detection, this paper introduces a feature-based image processing method. Feature extraction, dictionary learning and sparse representation are included in this method. Draped surface of wrinkling strip is suitable for SIFT technique extracting features of wrinkles. The extracted features will be collected by dictionary learning. For every original strip image, detecting wrinkles may be realised by assessing which dictionary its features incline to. At this stage, the assessment is implemented by sparse representation. The proposed method has been tested in a strip steel production line and the performance showed its applicability.

Keywords: strip wrinkling detection; SIFT feature extraction; dictionary learning; sparse representation; strip steel production; wrinkles.

DOI: 10.1504/IJWMC.2017.083051

International Journal of Wireless and Mobile Computing, 2017 Vol.12 No.1, pp.36 - 40

Received: 29 Apr 2016
Accepted: 06 Dec 2016

Published online: 18 Mar 2017 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article