A method for text-line segmentation for unconstrained Arabic and Persian handwritten text image

Abstract

One of the challenging parts of freestyle handwritten text documents recognition area is text line segmentation problem. Curvilinear text lines and small gaps between neighboring text lines present a challenge to algorithms developed for machine printed or hand-printed documents. In this paper, we propose a novel approach based on painting algorithm by dividing of a text image into number of vertical segments which is called striping. As Arabic and Persian scripts present a lot of dots, we considered historical available nastaliq scanned pages for experiments. Results show the proposed algorithm is robust to scale change, rotation, and noise. The proposed method may contribute significantly for the development of applications related to OCR.