06-11-2014, 03:53 PM
Abstracts: This project report concerned with implementation details of English character recognization system.The basic objective of the present work is to provide an efficient and reliable technique of recognization of printed characters.Three different type of techniques has been use for classification of characters 1)Thinning,2)Scaling and 3)Feature Extraction Rule. The whole project devide into three steps,1)Preprocessing 2)Classification 3) Post processing.In Preprocessing there is i)Noise Removal ii)linarization iii)Skew Detection and Correction iv) Thinning v)Scaling. After this we can classify the characters with the different methods.At last spellchecker is also available if something wrong into output because every ocr project is limit upto 99%,because there is lots of noise adds when we scan image from scanner.I try my best to reduce time and more accuracy to successful output generation.