Title :
A hybrid approach to recognize handwritten alphanumeric characters
Author :
Mandalia, A.D. ; Pandya, A.S. ; Sudhakar, R.
Author_Institution :
Florida Atlantic Univ., Boca Raton, FL, USA
Abstract :
The design characteristics of a hybrid approach involving expert systems and neural networks to recognize isolated handwritten alphanumeric characters is presented. The optical character recognition (OCR) system was designed to recognize the wide variation in writing style of alphanumeric characters consisting of uppercase characters, lower case characters, and numerals, a total of 62 characters. Issues concerning the performance and speed of the algorithms of the OCR system are addressed since the total character set is of significant size. The overall OCR system architecture consists of subsystems which were designed with considerations for hardware implementation. The key subsystems utilized the Hough transform for feature extraction, and neural networks and Dempster-Shafer theory for classification
Keywords :
Hough transforms; expert systems; feature extraction; neural nets; optical character recognition; Dempster-Shafer theory; Hough transform; character set; classification; expert systems; feature extraction; handwritten alphanumeric characters; hybrid approach; lower case characters; neural networks; optical character recognition; uppercase characters; Character recognition; Expert systems; Feature extraction; Handwriting recognition; Hardware; Neural networks; Optical character recognition software; Optical computing; Optical design; Writing;
Conference_Titel :
Systems, Man and Cybernetics, 1992., IEEE International Conference on
Conference_Location :
Chicago, IL
Print_ISBN :
0-7803-0720-8
DOI :
10.1109/ICSMC.1992.271542
