Optimum Illumination Mechanism for Steel Plate Inspection Using Wavelet Synthetic Image and Entropy-Based Discriminant Measure

Demand for both quality steel-plate products and the quality inspection of steel plates is increasing. In order to guarantee the quality of steel plates, consistent automatic inspection is required, but there is a lack of research on illumination mechanisms for automatic inspection. The steel plate has low contrast, non uniformity, and featureless. To configure an illumination mechanism to cope with this problem, a discrete wavelet transform is used to decompose the image into multiple resolution regions to improve the contrast of the image and highlight features. In order to select the optimum illumination mechanism, the discriminant measure applying the entropy function is applied to the numerical analysis of the regions. The experimental results show that the BI&CISUBW/SUB illumination mechanism exhibits the highest discrimination, so BI&CISUBW/SUB is selected as the optimum illumination mechanism.