A Fuzzy Appraoch To Handwirtten Devanagari Character Recognition. Single Stage,Multistage & Hybride Feature Extraction Techniques

      Аннотация: The recognition of handwritten Devanagari character recognition is a vital research area in the domain of pattern recognition. Various approaches have been involved in the past decades and intensive research is still being carried out at present. Devanagari is an Indo Aryan language spoken by about 71 million people mainly in the Maharashtra, state of India. This book presents a comprehensive study on handwritten Devanagari character recognition. The proposed work based on hybrid feature extraction techniques which includes Moment invariant, affine moment invariant techniques and combination of both i.e. MI & AMI along with Box method and various morphological operations. Extracted feature from each sample proceed for classification using Fuzzy logic. The proposed techniques are independent of size, slant, orientation, translation & other variation in Devanagari handwritten character. This book may be very useful for researcher & academician working in the domain of handwritten character recognition.

Ключевые слова: Single stage, Multi stage & hybrid feature extraction techniques, Guassian Membership Function