We propose in this paper a recognition system of Arabic hand-written words issued from literal amounts of Arabic checks. This system is based on the idea of the PERCEPTRO system developed by M. Côté for Latin word recognition. It is a specific NN, named Transparent Neural Network (TNN), combining a global and a local vision modeling (GVM -LVM) of the word. In the forward propagation movement, the former (GVM) proposes a list of structural features characterizing the presence of some letters in the word. GVM proposes a list of possible letters and words containing these characteristics. Then, in the back-propagation movement, these letters are confirmed or not according to their proximity with corresponding printed letters. The correspondence between the letter shapes and the corresponding printed letters is performed by LVM using the correspondence of their Fourier descriptors (FD), playing the role of a letter shape normalizer.
The choice of relevant features is very decisive in handwriting recognition rate. Our aim is to present some useful structural and statistical features and see their degree of variability. In this paper, we start with a description of the variability of the Arabic handwriting and the way how to reduce it. Four kinds of feature sets used by our handwriting systems are then presented evaluated and discussed. The comparison is carried on a database of images from IFN/ENIT databases. The Neural Network Multilayer perceptrons is our method of classification. A contrastive study of these primitives is done according to recognition their time and memory consuming and their variability degree.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.