A Novel Segmentation Method for Overlapping Cells

The clustering phenomenon often appears in the cell image auto-reading system, some cells overlap or touch together to form a big area. It is necessary to design an effective algorithm to separate the clustering cells into single ones. In order to realize the cell´s parameter measurement and get a right analysis conclusion, firstly overlapping cells should be exactly segmented. This paper presents an effective segmentation algorithm: based on the combination of arc-chord ratio and area of cells image, firstly overlap cells should be separated from normal ones by area; secondly, the border of overlapping cells should be traced by Freeman-code; last the max arc-chord ratio can be solved by Freeman-code, at the same time separation points can be solved. In practice, this algorithm is proved to be an effective and reliable method, and both the segmentation speed and repeatability meet the medical clinic need, and the analysis conclusion accords with clinic diagnoses