Network Approach based Hindi Numeral Recognition

Abstract

Handwriting has kept on persevering as a methods for correspondence and recording data in everyday life even with the presentation of new advancements. The steady improvement of PC apparatuses prompt the necessity of less demanding interface between the man and the PC. Written by hand character acknowledgment may for example be connected to Postal division acknowledgment, programmed printed frame securing, or checks perusing. The significance to these applications has prompted extraordinary research for quite a while in the field of disconnected manually written character acknowledgment. 'Hindi' the national dialect of India (written in Devanagri content) is world's third most prevalent dialect after Chinese and English. Hindi manually written character acknowledgment has got parcel of utilization in various fields like postal address perusing, checks perusing electronically. Acknowledgment of written by hand Hindi characters by PC machine is convoluted errand when contrasted with composed characters, which can be effortlessly perceived by the PC. This paper exhibits a plan to perceive hindi number numeral with the assistance of neural network.