Segmentation of Connected Components in Arabic Handwritten Documents

We propose a segmentation method to separate between touching and overlapping connected components in Handwritten Arabic documents. The segmentation requires a recognition and transformation stages in which we use shape context and thin plate spline to find, for a connected component the most similar template within stored models and estimate the aligning transform. Then, we segment the touching components using the central points of the identified model's parts. The proposed segmentation method seems to be also efficient in multi-touching cases of connected components. Tests are performed on a large database of connections and compared with a number of known techniques in the literature. The results prove the efficiency of the proposed method.