Fourier Transformation and Invariant Moments applied to Tifinagh Characters Recognition

Abstract

Optical Character Recognition OCR is a tool that aims to provide opportunities for computers to read characters without human intervention. The objective of OCR is characterization of a character by invariant descriptors in translation, rotation and scaling. In this paper, the OCR developed use invariant moments and Fourier transform in extraction phase. In the recognition phase, dynamic programming and neural network are adopted. All tests are applied on Tifinaghe printed characters.