Hybrid Arabic Handwritten Character Recognition Using PCA and ANFIS

In this paper we will present a two phase method for isolated Arabic handwritten character recognition system. The proposed system is a hybrid system that uses the principal component analysis (PCA) feature technique and neuro-fuzzy classifier. The Adaptive Neural Network Fuzzy Inference System (ANFIS)were used at all levels of the character recognition stages with different learning algorithms and nonlinear outputs. The proposed system is applied to the Sudan University Sciences and Technology Arabic Recognition Group (SUST-ARG) data set. In this paper the work was divided into two stages. In the first stage, the system was applied to 34 Arabic characters and achieved 96.2% recognition rate for the tested data set. In the second stage, a private classifier for each group was created to recognize and classify the characters within a group which achieved 99.5% recognition rate for the tested data set.