Abstract—In this paper, we are focused on characters recognition, for this we present a comparison between the Krawtchouk Invariant Moment (KIM) and the Pseudo Zernike Invariant Moment (PZIM) for the recognition of printed Arabic characters (translated, rotated and contaminated by noise). In the preprocessing phase, we use the thresholding technique, and in the learning-classification phases, we use the supports vectors machines (SVM).The simulation results demonstrates that the KIM method gives more significant results that the PZIM for each Arabic character.

Index Terms—recognition, printed arabic characters, krawtchouk invariant moments, pseudo zernike invariant moments, support vectors machines

Cite: Rachid Salouan, Said Safi, and Belaid Bouikhalene, "A Comparative Study between the Pseudo Zernike and Krawtchouk Invariants Moments for Printed Arabic Characters Recognition," Journal of Emerging Technologies in Web Intelligence, Vol. 6, No. 1, pp. 89-93, February 2014. doi:10.4304/jetwi.6.1.89-93