DESIGN OF A SEPARATING ALGORITHM FOR OVERLAPPING CELL IMAGES

The overlapping phenomenon (also called clustering) often appears in actual cell images, i.e., there are some cells clustering together to form a big area. This phenomenon causes difficulty for consequent process, so it's necessary to design an effective algorithm to separate the clustering cells into single cells. In this paper a feasible method for binary images is proposed based on mathematical morphology. First, an erosion operator is applied to determine the number of single cells from clustering cells, and then based on the erosion result, the separating points on the original binary image are searched and these separating points are used to separate the clustering cells into single cells. The design and implementation of algorithm are detailed in this paper, and the experimental results are given also.