Detection of scratch defects on slab surface

Non-destructive automatic inspection systems are in high demand in the steel industry, for controlling the quality and increasing the productivity of steel products. In this study, we developed an algorithm for detecting scratch defects on slabs by using image-processing techniques. To reduce the effect of lighting condition and amplify the features of the defects, we used a Gabor filter and carried out edge-pair detection. The histogram-based and gradient texture features, and support vector machine are used to detect the scratch defects. Experimental results proved the effectiveness of the proposed algorithm.