Recognition of Handwritten Odia Numerals Using Artificial Intelligence Techniques
Pradeepta Kumar Sarangi, Pervez Ahmed · 2013
Creation of an accurate and efficient recognition system for unconstrained handwritten characters has been a subject of intensive research. Till date, no technique is available that can recognize with 100% accuracy for all type of data sets that may be observed in real-life. That is why research is still on for finding a robust technique capable of dealing with natural variations in handwritten characters. In this paper, the suitability of Artificial Neural Network (ANN) model with back-propagation learning has been studied in recognizing the handwritten numerals of Odia language (an Indian regional language). LU decomposition of matrix has been used to extract the feature vectors from the character image.