Cell Segmentation from Cellular Image

To understand the cell movement and cell behavior into different parts of organs in human or animal body, it is necessary to study the cells in culture medium. Fluorescence microscopy is an emerging tool for acquiring this cellular images. The large number of cellular images produced by fluorescence microscopy is unmanageable for human to analyze them manually. Thus, cellular image segmentation is a primary requirement for higher level analysis of medical diagnosis and research. In this paper, a fully automatic method for segmentation of cells from fluorescent microscopy images is proposed. The method first mark the probable foreground and background seeds planted on the image. Then it applies a popular segmentation method, namely watershed segmentation, based on these seeds. The result is further refined based on the gradient value along the initial segmented lines and then again performing watershed transform on the distance transformed image of the previously founded result. The experimental results shows that this approach for segmenting cell images is both fast and robust.