A combined method of segmentation for connected handwritten on palm leaf manuscripts

Character segmentation of handwriting of ancient manuscript is a crucial step in Optical Character Recognition (OCR) system. The segmentation of connected components is one of the factors that affects the performance of the recognition system. In order to improve the efficiency of OCR systems, segmentation of touching characters is a significant task. This paper proposes a combined method for the handling of connected handwritten segmentation of Thai-Noi script on palm leaf manuscripts. This combined process of handwritten segmentation estimates points of connected positions from two methods: (1) points of foreground detection: foreground skeleton is established first, the path and junction are then detected using Hough transform and histogram of projection profile; and (2) points of background detection: background skeleton is extracted, the top and bottom paths are then detected by using Hough transform. The proposed method has been implemented and experimental results show that the combined method obtained better segmentation of handwritten documents on palm leaves.