摘要
采用一种改进的一次通过的并行细化算法对一般细化算法的不足进行了改进，细化后的文字骨架较好地反映了汉字的结构，适合于手写汉字的笔段提取，在细化的基础上，又提出了一种笔段提取的方法，该方法以细化骨架上的特征点为依据，通过对细化骨架的象素点跟踪获得初步的结果，再进行笔段合并处理．
An improved 'one - pass parallel thinning' algorithm is introduced to overcome the shortags of general thirming algorithm. The skeleton obtained fiom thirining may reflect the structure of original Chinese character in a better way and is suitable for extiacting strokes of Chinese characters. On the basis of thirining, a new method of stroke extraction is proposed. Through tracing the pixel points along the skeleton as well as taking the feature points of the skeleton into account the preliminary stroke segments may be obtained at first. Then the stroke combination process may finally be completed.
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
1999年第5期107-110,共4页
Journal of Harbin Institute of Technology
基金
国家863计划项目!863-306-63-05-3
关键词
手写体汉字识别
笔段提取
细化算法
汉字识别
off- line handwritten Chinese character recognition
stroke extraction
thinning
