BP神经网络在图像字符识别中的改进和应用  

Improvement and application of BP neural network in image character recognition

作　　者：张永宏[1] 吴鑫[1] 

机构地区：[1]南京信息工程大学信息与控制学院,南京210044

出　　处：《南京信息工程大学学报（自然科学版）》2012年第6期526-529,共4页Journal of Nanjing University of Information Science & Technology（Natural Science Edition）

基　　金：教育部留学回国人员启动基金(2010609);江苏省"六大人才"高峰资助项目(2010-JXQC-132)

摘　　要：传统BP神经网络算法虽然具有良好的学习能力和容错能力,但是收敛速度慢,易陷入局部极小点等缺点制约了它的进一步发展和应用.针对这些不足,采用自适应学习率结合附加动量因子的方法可以有效缩短训练时间,加快收敛速度,同时抑制寻优算法陷入局部极小点.将该算法应用于图像字符识别系统中,通过一系列实验优化系统参数之后给出系统识别结果,表明该系统识别具有较高的准确性和鲁棒性.The traditional BP neural network algorithm is good in learning ability and fault tolerance,while its dis- advantages such as slow convergence rate and easily falling into local minimum restrict its further development and application. An improved BP algorithm with self-adaptive learning rate and additional momentum factors can effec- tively reduce the training time, speed up the convergence rate and inhibit the possibility of falling into a local mini- mum. The improved algorithm is applied to the image character recognition system. The influences of model parame- ters on performance of BP neural network are analyzed, and the recognition results are given after a series of parame- ter optimization. The experimental results show that the improved BP neural network can recognize image characters with high accuracy and robustness.

关 键 词：字符识别 BP神经网络 动量因子 自适应学习率 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]