Nafiseh Salehian
MohamadReza Yazdchi
AliReza Karimian

This paper introduces a complete system for recognition of Farsi Nastaaligh handwritten words using Neural Networks. In preprocessing stage, after connected component specification, new algorithms are applied to find and eliminate ascenders, descenders, dots, and other secondary strokes from the original image. Then by using a segmentation algorithm based on analyzing upper and under contours, the word is segmented to a series of sub-words and their arrangement (Right to Left) is defined. Eight features, including three Fourier descriptors and five structural and discrete features, are applied to represent symbols in the feature space. Recognition is based on using a Feed Forward Back Propagation Network. The probable mistakes in recognition of sub-words will be corrected  by using a search algorithm in dictionary of system. Experiments on a sample of ۳۲۰ words show a suitable performance (%۹۷ correct recognition) of the system.

Segmentation, en, Neural network, word recognition