The recognition of electronic component's printed marks based on rough set

Rough set (RS) developed by Professor Z. Pawlak in 1982 has been applied successfully in such fields as machine learning, data mining, intelligent data analyzing, control algorithm acquiring, etc. In our paper, we apply this theory to recognize characters printed on the surface of electronic component. According to the properties of these characters, we propose a new method based on mathematical morphology to extract feature. Here, we propose a new algorithm to generate rules from decision table, and classify character based on the prior probability derived from such rule set In order to reduce the complexity further, we design a flow to pick train samples. The result of experiment shows our methods are useful and efficiency.