A Standardized Frame work for Handwritten and Printed Kannada Numeral Recognition and Translation using Probabilistic Neural Networks
N. Shobha Rani, P Ashwini · 2014
Numeral recognition is considered to be very prominent in most of the Character recognition researches. With respect to applications like number plate recognition and document processing the numerals are composed as a part of number plate images/application form type document images. This paper mainly focuses on eliminating language barriers that may arise while comprehending the regional language numerals by a non-regional user at the time of number plate recognition or other application form type document processing with special reference to Karnataka state. An algorithm is devised by incorporating the capabilities of functionalities of fast Fourier features and probabilistic Neural networks to recognize and convert the handwritten and printed Kannada numerals to Roman numerals.