Odia Character Recognition Using Curvelet Transform with DWT Feature Extraction

摘要

Developing of OCR for Indian scripts is an increasing area of research. This research work is an attempt to make an OCR system for Odia characters and numbers which is official language of ODISHA. In this work we proposed a new approach for Odia character recognition using digital curvelet transform. It presents characters and numbers recognition scheme for Odia characters. The curvelet transform although heavily utilized in various areas of image processing. The curvelet coefficients of a unique picture just as its morphologically adjusted renditions are utilized to KNN classifiers. The yield estimations of these classifiers are combined utilizing a straightforward dominant part casting a ballot plan to touch base at an official choice utilizing one of a kind structure of some character we have discovered better outcome as correlation with different techniques.