Title :
Optical character recognition for Sinhala language
Author :
Rimas, Mohamad ; Thilakumara, R.P. ; Koswatta, Priyaranga
Author_Institution :
Dept. Electr. & Comput. Eng., Sri Lanka Inst. of Inf. Technol., Malabe, Sri Lanka
Abstract :
A rotationally invariant optical character recognition system for Sinhala language is developed using Two Dimensional Fourier Transform and Artificial Neural Networks. Sinhala characters of different fonts and font sizes are recognized with over 85% recognition accuracy. Segmentation method based on histogram used in this system gives segmentation accuracy over 70% for complex Sinhala characters.
Keywords :
Fourier transforms; natural languages; neural nets; optical character recognition; Sinhala language; artificial neural networks; optical character recognition; segmentation method; two-dimensional Fourier transform; Character recognition; Feature extraction; Histograms; Image segmentation; Neural networks; Optical character recognition software; Optical imaging; Neural network; Sinhala;
Conference_Titel :
Global Humanitarian Technology Conference: South Asia Satellite (GHTC-SAS), 2013 IEEE
Conference_Location :
Trivandrum
Print_ISBN :
978-1-4799-1094-6
DOI :
10.1109/GHTC-SAS.2013.6629906
