Title :
Recognition of real-life character samples using a structural variation and degradation model
Author :
Khan, Nadeem A. ; Hegt, Hans A.
Author_Institution :
Dept. of Electr. Eng., Eindhoven Univ. of Technol., Netherlands
Abstract :
A character recognition scheme based on structural analysis is presented to deal with unconstrained handwritten characters and degraded samples of multi-font machine-printed characters. It is based on building simplified easy-to-construct shape models representing character classes. The scheme is based on employing a general variation and degradation model to match character samples with these models. Experimental results are presented as proof-of-concept
Keywords :
character sets; handwritten character recognition; image recognition; text analysis; character classes; character recognition scheme; degradation model; degraded samples; multi-font machine-printed characters; proof-of-concept; real-life character sample recognition; simplified easy-to-construct shape models; structural analysis; structural variation; unconstrained handwritten characters; Buildings; Character recognition; Degradation; Handwriting recognition; Prototypes; Robustness; Shape; Skeleton; Solid modeling; Text recognition;
Conference_Titel :
Document Analysis and Recognition, 1999. ICDAR '99. Proceedings of the Fifth International Conference on
Conference_Location :
Bangalore
Print_ISBN :
0-7695-0318-7
DOI :
10.1109/ICDAR.1999.791765
