Title :
BP alogrithm in pattern recognition of glass defects
Author_Institution :
Dept. of Inf. Eng., Environ. Manage. Coll. of China, Qinhuangdao, China
Abstract :
According to the characteristics of glass defects, through analyzing the advantages and disadvantages of the traditional BP algorithm, an improved BP neural network recognition algorithm is applied to the glass defect classification and character recognition. Experimental results show that compared with traditional BP recognition algorithm, convergence speed of the algorithm is fast and the identification of false positives is low.
Keywords :
automatic optical inspection; backpropagation; character recognition; convergence; glass; image classification; neural nets; production engineering computing; character recognition; convergence speed; false positive identification; glass defect classification; improved BP neural network recognition algorithm; pattern recognition; Algorithm design and analysis; Approximation algorithms; Classification algorithms; Glass; Neurons; Tin; Training;
Conference_Titel :
Advanced Computational Intelligence (ICACI), 2012 IEEE Fifth International Conference on
Conference_Location :
Nanjing
Print_ISBN :
978-1-4673-1743-6
DOI :
10.1109/ICACI.2012.6463147
