形态学运算在字符识别中的应用  

Application of Mathematical Morphologic Process in Character Recognition

作　　者：杨智勇[1] 周定康[1] 周琪云[1] 

机构地区：[1]江西师范大学计算机应用研究所,江西南昌330027

出　　处：《计算机应用研究》2004年第6期99-100,103,共3页Application Research of Computers

摘　　要：字符识别作为基本的图像识别技术 ,在计算机输入、交通和安防等领域都获得了广泛的应用。在进行字符识别前 ,需要先对图像进行灰度拉伸和阈值分割等一系列预处理以获得有效的目标字符特征。这些处理虽然可以增强目标字符的特征并滤除部分噪声 ,但同时也经常会造成目标字符的笔画发生断裂。由于待识别字符的笔画为较细的线性结构 ,采用常规的形态学开运算与闭运算对图像进行处理非但不能连接断裂的笔画 ,字符笔画还有可能因为细小而被抹去 ,使图像中的目标字符产生更严重的笔画断裂甚至缺失。讨论了数学形态学运算在字符识别技术中的应用 ,利用形态学闭运算的变形对含目标字符的图像进行增强处理 ,可在一定程度上有效消除字符的笔画断裂。实验数据表明 ,该方法能有效消除笔画断裂 ,进一步增强目标字符特征 ,提高字符识别正确率。As a basic recognize technique,character recognition is widely used in computer input,traffic,security keeping and other fields.Before characters are recognized,we must do a series of preprocessings to the original image in order to extract their characteristics,such as Gray Stretch,Threshold Transform and so on.These preprocessings can strengthen the characteristics of target characters and get rid of noises.However,they often make the strokes of the target characters broken.Because the structure of the strokes is very thin,normal morphologic opening or closing process will break the strokes more seriously.By reforming the morphologic closing process,this paper proposes a process to eliminate the breaks of target characters.Experiment results show that the process is effective.

关 键 词：形态学 闭运算 图像处理 字符识别 

分 类 号：TP391.43[自动化与计算机技术—计算机应用技术]