An incremental clustering based solder joint inspection algorithm for chip component

作者：Xie, Hongwei; Kuang, Yongcong; Ouyang, Gaofei; Zhang, Xianmin^*

来源：Journal of Computational Information Systems, 2011, 7(14): 5193-5200.

• Adaptive ability
• Automatic optical inspection
• Chip component
• Data sets
• False alarm rate
• False alarms
• Incremental clustering
• Incremental clustering algorithm
• Production environments
• Recognition rates
• Repair stations
• Representative sample
• Representative samples
• Solder joint inspection
• Solder joints
• Training efficiency

In practical applications, the AOI (automatic optical inspection) system for solder joint inspection has a high false alarm rate due to the frequently changed production environment. To overcome the problem, an incremental clustering based solder joint inspection algorithm is proposed in this paper. In the system, the clustering algorithm is used for solder joint inspection, and an incremental clustering algorithm is used to adjust inspection parameters automatically according to the feedback from the repair station by training the false alarm samples. To improve the training efficiency of incremental clustering, only some representative samples of old data set are selected and trained with new samples together. To evaluate the performance of the proposed algorithm, various types of solder joints were inspected by an AOI system which integrates with the proposed algorithm. The inspection results have verified the effectiveness of this method in terms of speed and recognition rate.

• 出版日期2011-12
• 单位华南理工大学