Handwritten Arabic Words Recognition System Based on HOG and Gabor Filter Descriptors

Abstract

Automatic recognition of handwritten Arabic words is a research area opens to a large number of industrial applications. However, the solution to the problem of cursive handwriting recognition still laborious because of the complexity of the morphology of Arabic script. In our work, we study and implement an offline handwritten word recognition system using the IFN / ENIT dataset of Arabic words representing the names of Tunisian cities. We have used two types of descriptors, the oriented gradient histogram (HOG) and Gabor filter for the extraction of features and as a classifier the kNN. Experiments are performed on samples extracted from the IFN/ENIT database. The results highlight the reliability of the kNN classifier for handwritten Arabic word recognition.