Defect detection algorithm for corner cracks in steel billet using discrete wavelet transform

The detection of defects on steel surfaces is an important quality control step in the manufacturing of steel products. In this study, we developed an algorithm to detect defects such as corner cracks in steel billets through the use of image processing techniques. The surface of a billet is partially covered with scales composed of oxidized substance. Because these scales visually similar to corner cracks, and because the surface features of billet images vary under different lighting conditions, defect detection can be difficult. To minimize the influence of scales and to improve the accuracy of detection, a new detection method based on discrete wavelet transform is proposed. Also, to distinguish corner cracks from pseudo defects, we investigated the morphological features of corner cracks.