Machine vision for defect detection on the air bearing surface

Air Bearing Surface (ABS) is a part on HDD responsible to control the flying height during read/write data. The defect on ABS causes poor performance, therefore it is necessary to verify the part before assembly to the final products. This paper presents a practical approach for machine vision to detect defect on the air bearing surface (ABS) of a head gimble assembly. Detail design on imaging system and contamination detection is provided. The proposed system is designed and implemented to detect the defect on ABS using image segmentation techniques with generation of block matrices. To determine defect in paralleled processing architecture, sub blocks and sub-ROI are allocated and employed. The results show the effectiveness of the designed system which can be deployed in the industrial assembly lines.