PCA+4点算法在手写体数字特征识别中的应用  

The Handwritten Digits Feature Recognition Based on the PCA+4 Points Algorithm

作　　者：张小红[1] 陈贞忠[2] 

机构地区：[1]河南财政税务高等专科学校信息工程系,郑州451464 [2]新乡学院数学系,河南新乡453000

出　　处：《河南师范大学学报（自然科学版）》2011年第2期160-162,共3页Journal of Henan Normal University(Natural Science Edition)

基　　金：河南省重点科技攻关项目(092102210149)

摘　　要：提出一种改进手写字体特征的提取方法:将传统的PCA特征方法与13点特征方法进行综合,得到一种PCA+4点的特征提取算法,然后通过BP神经网络进行训练识别.实验仿真表明这种改进的方法比PCA特征提取及13点特征提取的识别率高,特别在手写变化大、手写速度快等方面优势更加明显.The artificial neural network has been widely applied in the fields of information processing,pattern recognition and intelligent control,performing good intelligence characteristics particularly in the aspects of image recognition,speech recognition,memory,forecast and optimization.The research has proposed an improved method of feature extraction font by using artificial neural network,which has mixed the traditional PCA features with 13 points features,and has got a PCA+4 points of feature extraction algorithm.Through experiments emulation,the recognition rate of the improved method is higher than the methods of PAC feature extraction and 13 points feature extraction,the handwritten change is big in particular,and the quick handwritten speed is more obvious.Thus,this method has certain practical value in the system of font recognition.

关 键 词：PCA+4点 特征识别 细化 神经网络 

分 类 号：N55[自然科学总论] G658.3[文化科学—教育学]