A fuzzy SVM for intelligent diagnosis of solder bumps using scanning acoustic microscopy

Flip chip technology has been used extensively over the past decades. In this paper, an intelligent diagnosis system for solder bumps based on the scanning acoustic microscopy (SAM) has been developed, and a fuzzy support vector machine (F-SVM) algorithm is investigated for defect classification. An ultrasonic transducer of 230MHz is used to capture the image of the flip chip. Solder bumps are then segmented according to the gradient matrix of the SAM image, and the statistical features are extracted and adopted for bump classification. The experimental results show that the F-SVM algorithm reaches a high recognition accuracy, and the intelligent diagnosis system is effective for defect inspection of solder bumps in high density packaging.