An Optimized Approach for Handwritten Arabic Character Recognition based on the SVM Classifier

Optical Character Recognition (OCR) is an essential technology, capable of addressing complex challenges while simplifying numerous human activities. Although it emerged in the 1970s with various solutions, these efforts primarily focused on Latin-based languages, leaving other writing systems, such as Arabic, largely underexplored. In this context, this study proposes an innovative offline Arabic handwriting recognition system based on a structural segmentation method combined with the use of Support Vector Machines (SVM) for character classification. An in-depth review of different character segmentation methods was followed by an in-depth analysis of the OCR field. This study examined the challenges associated with normalization, a recurring issue in the processing of handwritten scripts. Finally, after comparing the unique characteristics of Arabic handwritten characters with existing segmentation techniques, an approach was developed based on a segmentation algorithm to improve the accuracy and efficiency of the recognition process.