论文部分内容阅读
本文实现了一种基于机器学习的手写汉字识别方法.针对手写汉字的特点,选择并提取了横竖笔划特征、周边特征、结构划分特征、分区特征点、黑点重量等作为分类特征.在分类策略中采取了先粗分类后细分类的多级分类方法.并将决策树算法ID3成功地应用到分类策略中,在识别时中利用决策树引导特征提取,减少了特征提取的数量,从而大大提高了识别速度.
In this paper, a handwritten Chinese character recognition method based on machine learning is implemented.According to the characteristics of handwritten Chinese characters, the vertical and horizontal strokes, peripheral features, structural features, partition feature points and black point weights are selected and extracted as classification features.In the classification strategy , A multi-level classification method is proposed, in which the rough classification and the fine classification are adopted, and the decision tree algorithm ID3 is successfully applied to the classification strategy, and the decision tree is used to guide the feature extraction during recognition, thereby reducing the number of feature extraction and greatly increasing Recognition speed.