열간 슬라브 표면 결함 검출 알고리즘

윤종필, 박창현, 배호문 et al. · 제어로봇시스템학회 국내학술대회 논문집 · 2010

This paper presents a new defect detection algorithm for scale-covered steel slab surfaces. The slab surfaces are covered with scales, which are formed due to the deposition of unavoidable oxidized substances produced during manufacturing. The presence of scale makes it very difficult to accurately defect detection. In this paper, we focus on the burr defects among various surface defects. In order to facilitate the development of the detection algorithm, burr defects are divided into two classes according to the defect shapes namely, projection burr and residual burr. The binarization with threshold, horizontal projection profile, and sobel filter are used to detect projection burr. On the other hand, discrete wavelet transform, texture analysis based on histogram and support vector machine are applied to detect residual burr. Finally, the experimental results conducted on slab surface images from actual steel production line show the effectiveness of the proposed algorithm.

Cited by 1 works.

Authors