Application of Support Vector Machines for Recognition of Handwritten Arabic/Persian Digits

A new method for recognition of isolated handwritten Arabic/Persian digits is presented. This method is based on Support Vector Machines (SVMs), and a new approach of feature extraction. Each digit is considered from four different views, and from each view 16 features are extracted and combined to obtain 64 features. Using these features, multiple SVM classifiers are trained to separate different classes of digits. CENPARMI Indian (Arabic/Persian) handwritten digit database is used for training and testing of SVM classifiers. Based on this database, differences between Arabic and Persian digits in digit recognition are shown. This database provides 7390 samples for training and 3035 samples for testing from the real life samples. Experiments show that the proposed features can provide a very good recognition result using Support Vector Machines at a recognition rate 94.14%, compared with 91.25% obtained by MLP neural network classifier using the same features and test set.

کلیدواژه ها:

Optical Character Recognition (OCR) ، Feature Extraction ، Machine Learning ، Support Vector Machine (SVM) ، Multiple Support Vector Classifiers ، MLP Neural Network