Microscope Image Processing

15.6.1 Cell Tracking

Possibly the most extensively studied approach to image segmentation in recent

years is the use of level-set methods [63]. These methods have also been explored

for cell segmentation and tracking [56, 57, 96], with promising results. Most

classical model–based approaches involve cumbersome explicit representations

of objects by marker points and parametric contours or surfaces [62, 97].

In contrast, level-set methods conveniently define object boundaries, in an

implicit way, as the zero level set of a scalar function, denoted by w(.) here.

This level-set function is defined such that w(x) > 0 when x lies inside the object,

w(x) <