Solder Joint Inspection Method for Chip Component Using Improved AdaBoost and Decision Tree

To improve the performance of automatic optical inspection (AOI), a new inspection method for chip component of mounted components on printed circuit boards is developed. In this paper, the inspection procedure is divided into training stage and test stage. In the training stage, first, the solder joint is divided into several sub-regions according to priori knowledge, second, various features in every sub-region are extracted, then, for every sub-region the optimal features are selected with an improved AdaBoost by evaluating their classification ability and independency, finally, the classifier for every sub-region is established with the selected features by training a number of samples. In the test stage, after image acquirement the inspection of a solder joint consists of region division, critical features extraction, classification of sub-regions, and defect diagnosis. The former three steps are executed based on the training results, and in the last step a new defect binary decision tree based on classification and regression tree is used to determine the final defect type. To evaluate the performance of the proposed method, various types of solder joints were inspected by an AOI system which integrates with the proposed method. The inspection results have verified the effectiveness of this method in terms of speed and recognition rate.