Research of Classification System Based on Naive Bayes and MetaClass

Considering the problem that image defect´s fineness, complex shape, difficultly to extract feature, and easily effected by noise on PCB products machine vision inspection system, the paper presented defect identification classification algorithm based on Naive Bayes and MetaClass , which resolved the problem that fine and complex defect is difficult to classify. Regard Naive Bayes algorithm to construct the binary tree in multi class MetaClass classification algorithm. It resolved the problem that the structure of binary tree affected the accuracy of classifier, and upgraded defect classification accuracy finally. The experiments show that six defects discrimination of this method is up to 96.2%, higher than BP network´s best discrimination 92.3% and 81.7% by method based on region,which the training and inspecting time is few. Verified this method efficiency from theory and experiment, and it has great value for research and usage.