Title :
A study on recognition method of shortforms in Chinese handwritten shorthand
Author :
Xiao-Ping, Chen ; Tie-Cheng, Yu
Author_Institution :
Lab. 5, Acad. Sinica, Beijing, China
Abstract :
This article aims to describe a structural recognition method for the shortforms in Chinese handwritten shorthand (CHS). Further analysis for the shortforms of CHS shows that, although differences between the same shortform written by two persons exist, the initial consonant notations and rhyme notations constituting the shortforms written by two persons are comparable. Therefore, a structural approach based on the initial consonant notation and rhyme notation primitives is applied in the recognition of shortforms, which first segments a given shortform into a string of primitives, and then classifies each primitive, finally giving the recognition result of the whole shortform based on lexical knowledge. Test results show that this structural recognition approach is applicable to the recognition of shortforms, and the experimental results are encouraging
Keywords :
feature extraction; handwritten character recognition; Chinese handwritten shorthand; feature extraction; initial consonant notations; lexical knowledge; primitives; recognition method; rhyme notations; shortforms; structural recognition approach; structural recognition method; Acoustics; Automation; Character recognition; Elasticity; Handwriting recognition; Laboratories; Natural languages; Pattern recognition; Testing; Text recognition;
Conference_Titel :
Signal Processing Proceedings, 1998. ICSP '98. 1998 Fourth International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-4325-5
DOI :
10.1109/ICOSP.1998.770844