摘要
提出了一种新的基于高斯概率模型的字符识别算法,该算法根据模式识别的样本分布特征与高斯分布的一致性,构建了一个高斯概率模型。在模型中存储概率为P的训练样本,分类识别时,将测试样本与模型进行相关计算得出概率值,进行判断。结果表明,该算法识别速度快,准确率高,与其他字符识别算法(KNN)相比有更好的实用性。
A new character recognition algorithm based on Gaussian probability model is presented,which constructs a Gaussian probability model based on the sample distribution characteristics of pattern recognition and the uniformity of Gaussian distribution.Training samples with probability P are stored,and when the classification and recognition are carried out,the probability value is calculated to be judged by the test samples associated with the model.The results show that the algorithm has fast recognition speed,high accuracy,and has better practicality than other character recognition algorithm(KNN).
出处
《测控技术》
CSCD
2017年第3期33-36,共4页
Measurement & Control Technology
基金
河南理工大学博士基金资助项目(B2010-17)
河南省科技攻关项目(162102210228)
国家自然科学基金项目(U1504623)
关键词
字符识别
高斯分布
概率模型
KNN
character recognition
Gaussian distribution
probability model
KNN