BP神经网络算法的改进及其在手写体汉字识别中的应用

The Improves on the Standard BP Algorithm and Its Use in the Field of the Off-Line Handwritten Chinese Character Recognition

分析BP算法的基本原理,指出BP算法具有收敛速度慢、易陷入局部极小点等缺陷以及这些缺陷产生的根源.针对这些缺陷,通过在标准BP算法中引入变步长法、加动量项法等几种方法来优化BP算法.应用实例利用MATLAB软件对标准BP算法及其改进的算法进行语言编程、仿真.实验结果表明,这些方法有效地提高了BP算法的收敛性,避免陷入局部最小点.同时,将改进的BP神经网络算法应用于脱机手写体汉字识别系统的实现,实验表明系统较好地回避了汉字结构复杂、变形难以预测等问题,提高了识别率.

Basic principle of BP algorithm is analyzed firstly. Then some defects such as slow convergence rate and getting into local minimum in BP Algorithm are pointed out, and the root of the defects is presented. Finally, in view of these limitations, several methods such as genetic algo rithm and simulated annealing algorithm etc. are led to optimize BP algorithm. At the same time given to show the differences between the standard BP algorithm aml the improved algorithm using MATLAB. Experiment results show that these methods increase effi-ciently the conver- gence performance of BP algorithm and avoid local minimum. A technological research on recognition of handwrit- ten Chinese characters based on these improved BP algorithms is summarized. The Chinese character recognition system is realized,as a result of the use of improved BP networks. Experimental results demonstrate that this system based on the improved BP algorithms can successfully overcome obstacles out of the complexity of the Chinese characters＇ structures,and of difficulties to forsee written forms＇ changes.