Title :
High performance embedded OCR algorithm for hand-held devices
Author_Institution :
Coll. of Math. & Comput. Sci., Jiangxi Sci. & Technol. Normal Univ., Nanchang, China
Abstract :
The Gabor feature and modified quadratic discriminant function (MQDF) have shown superior performance in character cognition. However, the huge storage of the parameters and great time-consuming make the methods not practical to be used in hand held device with embedded system. This paper presents an improving approach of Gabor feature extraction as well as simplified classifier of modified quadratic discriminant function (MQDF). Experiments testified that these methods are simple and effective in embedded systems.
Keywords :
Gabor filters; feature extraction; image classification; optical character recognition; Gabor feature extraction; character cognition; handheld devices; high performance embedded OCR algorithm; modified quadratic discriminant function; Character recognition; Classification algorithms; Embedded systems; Feature extraction; Gabor filters; Transforms;
Conference_Titel :
Audio Language and Image Processing (ICALIP), 2010 International Conference on
Conference_Location :
Shanghai
Print_ISBN :
978-1-4244-5856-1
DOI :
10.1109/ICALIP.2010.5684537
