A Review: Handwritten Character Recognition System

Abstract: This method presents an innovative technique to deals with the problem of offline handwritten character recognition of English character. This system is very useful to the people which have a lot of paper work on a daily basis. Handwritten character recognition systems provides the solution for how to convert handwritten words into computer readable format. Image enhancement and noise removal are done at the pre-processing phase. After enhancing the image, segmentation of the character will be done using different methods like Horizontal and Vertical Projection Profile method. Feature extraction will be done by using Scale Invariant Feature Transform (SIFT), chain codes, zoning etc. The Support Vector Machine (SVM), Artificial Neural Network (ANN), Euclidian distance methods will be used to recognize the characters. By following above mentioned processes handwritten article images will be converted into an editable text files.

Keywords: Feature extraction, handwritten character recognition, segmentation, SIFT, Support Vector Machine.