Handwritten Arabic Alphanumeric Character Recognition using BP and SOFM NeuralNetworks

AbstractThis paper presents results obtained by applying two neural networksmodels Backpropagation (BP), and Self-Organized Feature Map (SOFM) to a newapplication of handwritten Arabic alphanumeric character (HAAC) recognition. Anovel method for features extraction, based on a shadow projection is used. Bothnetworks are trained using Arabic character samples written by different people(learning set). They are required, after the learning is over, to recognize charactersout of the learning set. Evaluation of the recognition (classification) capability ofthe two models for 28 alphanumeric characters is achieved. Depending on theexperimental results, a comparison of both algorithms is done.