Stroke edge based binarization algorithm for the palm leaf manuscripts
Peng Ge, Pengfei Yu, Haiyan Li, Hongsong Li · 2017
A binarization algorithm based on strokes edge detection is presented which could binary the images of palm leaf manuscripts effectively with pretty performance in this paper. Firstly, the binarized contrast map is obtained by using an adaptive algorithm. Secondly, the stroke edge map is detected by combining the binarized contrast map with Canny edge map based on the bitwise AND operator. And then, get an exact stroke edge map by using a connected component labeling based denoising algorithm. Finally, distinguish the stroke pixels from the background pixels according to the gray values of the stroke edge pixels. The proposed algorithm is compared with some relevant algorithms and the experimental results demonstrate the superior performance of our work.