Segmentation of Blood Cell Micrograph Based on Geodesic Active Contour

To extract quantitative and shape information of blood cells from micrograph images, good segmentation of these images is essential. In this paper, a method Based on the geodesic active contour (GAC) is proposed for such task. The Gaussian filtering working for edge detection in GAC is replaced with wavelet-package filtering, which enable the GAC cope better with noise disturbance and finally lead to better location of cell boundaries. Also, The SUSAN operator and some morphologic operations, which are applied to search the initial contour close to object boundary, makes the GAC less influenced by initial condition. Finally, with the separation of connected cell regions existing in GAC outputs done, individual cell regions can be achieved. Experimental results show the advantage of the proposed method in accuracy.