A robust fully automatic scheme for general image segmentation

This work proposes a robust fully automatic segmentation scheme based on the modified edge-following technique. The entire scheme consists of four stages. In the first stage, global threshold is computed. This is followed by the second stage in which positions and directions of the initial points are determined. Local threshold is derived based on the histogram of gradients from the third stage. Finally, in the fourth stage, searching procedure is started from each initial point to obtain closed-loop contours. The whole process is fully automatic. This avoids the disadvantages of semi-automatic schemes such as manually selecting the initial contours and points. Additionally, the sensitivity to the selection of the threshold value from the watershed schemes can be dramatically improved. The proposed automatic scheme can reduce human errors and operating time tremendously, it is also more robust than the conventional segmentation schemes and applicable on various image and video applications.