Isolated Handwritten Arabic Character Recognition Using Freeman Chain Code and Tangent Line

Recognition of handwritten Arabic text is a difficult task since there are many challenges and obstacles that face any handwritten Arabic OCR system. Some of them include, but are not limited to: different handwriting styles, different characters that have similar contours, and the same character may have different forms according to its position in a sentence. Several approaches have been attempted to accurately recognize handwritten Arabic characters. However, the issue of the accuracy of Arabic OCR in handwritten text continues to be a dilemma. We will describe the general difficulties in handwritten Arabic language text, and propose a novel approach for identifying isolated handwritten Arabic characters using encoded Freeman chain code. We will also apply a novel approach of using change in tangents to classify characters. Several handwritten Arabic characters were trained and tested with our own dataset. The results showed the efficacy of our approach for recognizing isolated handwritten Arabic characters. The average accuracy rate of our method ranges from 92% to 97%.