Abstract
Electronics manufacturing and IC packaging & testing industries confront the tendencies of higher component density and Pb-free manufacturing process in the future. Traditional two-dimension automatic optical inspection machine is limited by several factors, such as angles of the light source and photographing, it is not able to fulfill the need in the future. Meanwhile, the technique of using x-ray image to reconstruct three-dimensional image to inspect the product gradually grab the attention of the industry. It is believed this approach will play an important role in the near future. Since computed tomography (CT) is developed, computer has a powerful arithmetic ability that can reconstruct the section images of structure, which is a great contribution to modern medical science. Nowadays, in order to apply this technique in manufacturing and inspection process, its basic principles and technologies must be carefully studied. In this dissertation, finding the best parameter of filtered back-projection (FBP) method by the correlation coefficient (CC) can rebuild the 3D BGA structure and these images can detect the deficiencies is existed or not.