A Contrast Adjustment Thresholding Method for Surface Defect Detection based on Mesoscopy

Titanium coated surfaces are prone to tiny defects such as very small cracks which are not easily observable by the naked eye or optical microscopy. In this study, two new thresholding methods, namely, Contrast Adjusted Otsu’s Method and Contrast Adjusted Median-Based Otsu’s Method, are proposed for automated defect detection system for titanium-coated aluminum surfaces. The two proposed methods were compared with four existing tresholding techniques in term of accuracy and speed of defect detections for images of 700, 900 and 1000 dpi obtained using high resolution scanning. Experimental results have shown that the contrast-adjusting methods proposed have performance similar to Minimum Error Thresholding and is generally better than Otsu’s Method.