A Comparative study of Arabic handwritten characters invariant feature
Résumé
—this paper is practically interested in the unchangeable feature of Arabic handwritten character. It presents results of comparative study achieved on certain features extraction techniques of handwritten character, based on Hough transform, Fourier transform, Wavelet transform and Gabor Filter. Obtained results show that Hough Transform and Gabor filter are insensible to the rotation and translation, Fourier Transform is sensible to the rotation but insensible to the translation, in contrast to Hough Transform and Gabor filter, Wavelets Transform is sensitive to the rotation as well as to the translation I. INTRODUCTION In order to improve the recognition rate of Arabic handwritten character, several techniques based on geometric correction or mathematic transform have been tested and applied on the input image of such writing. The purpose of all these techniques is to attempt to find the invariant features of Arabic handwritten characters. Unfortunately, all these attempts failed because, researchers didn't try to take advantages of each one of these techniques and then try to build an approach based on the cooperation of some (or all) of them. Consequently, we show in this paper the behavior of some well known mathematic transforms to decide in a next step on which of them can be used especially in a future Arabic OCR system based on the cooperation of several approaches and techniques.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...