Research on Dongba pictograph character recognition methods

摘要：

　　Dongba pictograph of China is presently accepted as the only pictograph in use,and Ancient Dongbaclassics are listed by the UNESCO as the “Memory of the World in urgent need of rescue and inheritance through modern information means.This paper argues the theory of Dongba pictograph character recognition based on the unique structural features of Dongba pictograph,and initially puts forward the implementation methods from image preprocessing stage to other processing stages like feature extraction,template matching,and neural network recognition,etc.and proves that these methods are feasible after the verification of experiments.

关键词： handwritten character recognition feature extraction template matching neural network

作者: Yang Meng Xu Xiaoli Wu Guoxin Wang Hongjun Ding Chunyan

作者单位: Beijing Key Laboratory on Measurement and Control of Mechanical & Electrical System,Beijing Informat MinzuUniversity of China,Beijing,100080,China

会议类型: 国际会议

会议名称: The Fifth International Symposium on Test Automation & Instrumentation(2014第五届国际测试自动化与仪器仪表学术会议)

会议地点: 北京

会议语种:英文

页码: 302-307

在线出版日期: 2014-09-23（万方平台首次上网日期，不代表论文的发表时间）