Automatic Detection And Counting Of Floating Cells In Medical Images

Based on image-processing techniques, the actual images of live cells from the Medical College of Jinan University are processed and analyzed in order to count automatically the live cells in this thesis. In the laboratory of the Medical College, the live cells is normally counted by a method called "cell-count boards", in which human's eyes are employed to count the live cells one by one. In this new system, we first focused on the characteristics of cell images. Then after applying image-processing techniques---using bilinear interpolating and fitting to the background of the cell images, we can count the live cells by the techniques of thresholding and contour tracking. The experimental results showed that our techniques are satisfactory as well as convenience.