基于小波变换的图像汉字笔划特征提取方法

Image Chinese character stroke feature extraction based on wavelet transform

借鉴仿生模式识别的认知观点,从汉字的构造机理和人类认识汉字的习惯角度出发,提出一种基于小波变换的图像汉字识别方法。制定了图像汉字笔划特征提取的具体规则,采用小波变换的方法对图像汉字边缘和笔划轮廓进行检测,通过有效提取图像汉字笔段信息,进行笔段合成,生成汉字或汉字的基本笔划。仿真实验结果表明,这种方法提高了图像汉字笔划特征提取的准确率和稳定性,对于印刷体和书写较规范的手写体图像汉字具有极高的识别率。

Based on the cognitive view of the bionic pattern recognition,from the structure mechanism of Chinese characters and human＇s habits of understanding it,this paper puts forward an image Chinese characters recognition method based on wavelet transform.The specific rules about image Chinese character strokes features extraction are developed.Image Chinese characters outline and strokes outline are detected with wavelet transformation.The effective strokes information is extracted and combined.Strokes or a Chinese character is generated.The simulation results prove this method can improve the accuracy and stability of the image Chinese characters stroke features extraction.It has higher recognition rate for print and normative handwritten image Chinese character.