Oriented Basic Image Features Column for isolated handwritten digit

Abstract

Several approaches for handwritten digits recognition are proposed an appearance feature-based approach. In this paper we process handwritten digit image without deskewing using oriented Basic Image Features (oBIF) Column scheme extracted from the complete image as well as from different regions of the image by applying a uniform grid sampling to the image. oBIF Column scheme is a very efficient feature descriptor for handwritten digits which is arise from variations in size, shape and slant. Moreover, 4th Nearest Neighbor (4-NN) has been employed as classifier which has better responses. The experimental study is conducted on MNIST dataset and 98.32% recognition rate has been achieved which is comparable with the state of the art.

DOI: 10.1145/3129186.3129189

8 Figures and Tables

Cite this paper

@inproceedings{Gattal2017OrientedBI, title={Oriented Basic Image Features Column for isolated handwritten digit}, author={Abdeljalil Gattal and Chawki Djeddi and Youcef Chibani and Imran Siddiqi}, booktitle={ICCES '17}, year={2017} }