METHOD FOR SEGMENTATION OF LOW CONTRAST CYTOLOGICAL IMAGES BASED ON THE ACTIVE CONTOUR MODEL

摘要

Within the problem of automated diagnostics of hematological diseases, the task of cell nuclei segmentation of cytological specimen images is considered. A two-level method for segmentation based on the active contour model is proposed. The main features of the method are: (a) the implementation of different strategies for active contour model design at coarse and precise levels of approximation, and (b) implementation of the modified Gaussian filter based on the heat equation with heat source or sink. Computing experiment has shown the efficiency of the proposed method.