Title :
The examination system of steel ball surface defect based on BP neural network
Author_Institution :
Inst. of Inf. Eng. of NIT, Nanchang, China
Abstract :
In view of the steel ball examination domain particularity, the more effective neural network structure has been obtained through the theoretical analysis and the massive experiments. An algorithm of steel ball surface examination has been proposed based on the BP neural network in this paper. The images pre-treatment should improve the examination rate. Finally the examination system of steel ball is proposed based on neural network.
Keywords :
automatic optical inspection; backpropagation; feature extraction; neural nets; steel; BP neural network; examination rate improvement; image pretreatment; steel ball surface defect examination system; surface defect;
Conference_Titel :
Information Science, Electronics and Electrical Engineering (ISEEE), 2014 International Conference on
Conference_Location :
Sapporo
Print_ISBN :
978-1-4799-3196-5
DOI :
10.1109/InfoSEEE.2014.6946286