摘要
针对图像中字符识别技术的应用需求,提出了轮廓跟踪算法的改进算法,在链码表示法基础上提出了轮廓表存储结构。在字符识别中根据轮廓表生成字符的轮廓特征数据序列,运用其相关值来度量字符的相似程度。实验结果表明提取特征少,识别速度快,准确率高,受字号的影响较小。
To fulfill the application requirement of character recognition in image, an improved algorithm for contour following algorithm was proposed and a contour list based on chain code was presented. In character recognition, the character lineament sequence was generated according to contour list and the character similarity was measured by its correlation value. The experimental results show that there are fewer extracted character features, faster recognition and higher degree of accuracy, and it is less affected by font size.
出处
《计算机应用》
CSCD
北大核心
2006年第10期2378-2379,2382,共3页
journal of Computer Applications
基金
陕西理工学院科研基金项目(SLG0330)
关键词
轮廓跟踪算法
轮廓表
字符识别
contour following algorithm
contour list
character recognition
