Persian Handwritten Word Recognition Using Zernike and Fourier-Mellin Moments
Kianoosh Bagheri Noaparast, Ali Broumandnia · 2009
This Article proposed a holistic method to recognize Farsi words. By using the Zernike moments, we managed to overcome sensitivity problem scale changes and rotation of words in the process of recognizing them. Novelty of this proposed method in which that we can extracting feature and recognizing word in a shorter time in comparison to other methods. Therefore, the suggested method could pave the way for conducting further relevant researches of pattern recognition. In order to segmentation of Farsi words, we apply connective components analysis. For recognize Farsi words, we used Elman neural network to train and test the system. A simulation result shows superiority of novel scheme over similar one, and improves recognition speed by about 12.5 times in average.