摘要
提出了一种基于游程特征的中英文本图像识别方法,用游程统计特征提供的图像信息作为图像模式识别的描述特征,在此基础上利用神经网络来训练分类器。实验结果表明,该方法的识别精度较高,具有一定的容错能力。
This paper presents a method for document image recognition based on run feature.The image represented by run feature provides the description for image pattern recognition.Based on it,Neural Network learning architecture is employed to learn to recognize the instances of the object class.Experimental results show that the method achieves high recognition accuracy,and it is perfect for fault tolerant.
出处
《计算机工程与应用》
CSCD
北大核心
2010年第7期194-196,共3页
Computer Engineering and Applications
关键词
图像识别
游程特征
倾斜度
image recognition
run feature
gradient
