Image Based Contamination Detection on Hard Disk Head Gimbal Assembly

Abstract

Contamination may appear in various stages of hard disk manufacturing including the head gimbal assembly. Currently, the detection of contamination requires manual intervention. An image based automatic contamination detection strategy is therefore presented. After a preprocessing step, the contamination detection algorithm first detects potential areas of contamination using circle detection. Then, each of the contamination contenders are classified as either a contamination or noncontamination using a set of specific rules. The algorithm has been tested on 1,050 head gimbal assembly images of which 313 depicted contaminations. Our preliminary results yields an accuracy of 73.8% with a false negative rate of 34.8% and a false positive rate of 23.6%. Future work includes finetuning the contamination classification rules.

Detection of micro contamination in hard disk drives using maximum likelihood estimation and angle detection

“…Yammen et al [4] explored the inspection of corrosions on pole tips at the end of an airbearing slider using area-based and contour-based features. In another image-based attempt [5] three steps were explored to identify contamination on the ABS. First, the input image was preprocessed to segment ABS sub-image from the original topview image of the HGA.…”

“…Finally, each of the contamination contenders were classified as being either a contamination or not by using angle measurements and an intensity threshold. The method proposed in [5] does not rely on template images. However, the angle measurement and intensity threshold caused many false detections with the low quality of the test images that were used.…”

“…The morphological skeleton operator is subsequently applied to the binarized image of the template image to obtain a skeleton of the ABS binary image. The circles detected in the previous step are traced on the skeleton image with a 50% larger radii (radiusnew = (6/4) • radiusold) to find points where the circumference of the detected circles crosses the skeleton [5]. Each circle is considered according to three cases depending on the number of cross-point on each circle; case:1 is number of cross-points = 0, case:2 is number of cross-points =1, and case:3 is number of cross-points ≥2.…”

“…The line between the cross C is denoted AC, and the line between the cr is denoted BC. If there are more than two cross than one angle is calculated and the largest angle is selected [5]. Example of angles between two cross points and center are shown in Figure 6.…”

“…This test suite comprised 1,050 images without contaminations, 500 images of these non-contamination images were used to calculate the likelihood and the prior probability, and 313 images with contaminations which were used to calculated the likelihood and the prior probability. The results were compared to those reported in [5] where angle measurement and intensity were used to identify contamination in ABS images. We also report the results when 1) the proposed method used the likelihood, the prior probability and angle measurement without template registration, 2) the proposed method used only the likelihood without template registration, and 3) the proposed method used the likelihood and angle detection method with template registration.…”