A Novel Approach for Cell Segmentation Based on Directional Information

Considering that medical images play a crucial role in today´s biomedical research, it is more and more important to extract the edges of cell images. However, there some phenomena of the overlapping and discontinuity of edges exist, so it is more difficult to solve the problems. In view of it, we have employed a novel scheme to integrate edge detection, edge thinning and edge linking by use of directional information. The scheme proposed in this paper performs significantly better than previous isolated methods (edge detector, edge filter and so on) so that we can obtain the integral edge information to further analyze the roundness, perimeter and number of cells. In a word, our scheme is focused on directional information from beginning to end. The experiments on some cell images indicate that our scheme achieves high quality and it is also applicable to similar images.