The Study and Implementation of Blood-cell Image Segmentation
Zhihong Feng · Computer Knowledge and Technology · 2009
A variety of strategies are presented on a given cell original image to separate cell-nucleolus and cell-plasm. We use three different image image segmentation methods (edge detection,best thresholding and watershed) to achieve image segmentation, and compare the advantage and disadvantage of there methods. Theoretical analysis and experimental results show that this approach is basically able to extract the nucleus, and maintain its shape, and the best thresholding method can achieve better results.