Image based contamination detection on hard disk head gimbal assembly

摘要

Contamination may appear in various stages ofhard disk manufacturing including the head gimbalassembly. Currently, the detection of contamination requiresmanual intervention. An image based automaticcontamination detection strategy is therefore presented.After a preprocessing step, the contamination detectionalgorithm first detects potential areas of contamination usingcircle detection. Then, each of the contamination contendersare classified as either a contamination or noncontaminationusing a set of specific rules. The algorithm hasbeen tested on 1,050 head gimbal assembly images of which313 depicted contaminations. Our preliminary results yieldsan accuracy of 73.8% with a false negative rate of 34.8% anda false positive rate of 23.6%. Future work includes finetuningthe contamination classification rules.