Abstract:
Offline text recognition is very important in a variety of applications, such as automatic sorting of mail, altering ancient documents, and historical document analysis. This research aims at the ensemble of different classifiers to improve Arabic handwritten word recognition. Features based on Histogram of Oriented Gradients and local binary patterns are computed to represent the handwritten words. Whereas Principal Component Analysis is used to reduce the space of the feature dimensionality of each of them. Each set of features is fed to Support Vector Machine classifier separately. Two independent classifiers are to be produced. A Bayesian method is applied to combine the output of the two classifiers. The proposed scheme was evaluated on the AHDB database of Arabic handwritten words. Combining the classifiers results in improved recognition rates, which, in some cases, outperform the state-of-the-art recognition systems. The proposed method has yielded significant achievements comparing with other studies.
