Robust Cell Image Segmentation Methods¹

E. Bengtsson, C. Wählby, and J. Lindblad

Center for Image Analysis, Uppsala University, Lägerhyddsv. 3, Uppsala, 75235 Sweden

e-mail: ewert@cb.uu.se

Abstract—Biomedical cell image analysis is one of the main application fields of computerized image analy-
sis. This paper outlines the field and the different analysis steps related to it. Relative advantages of different
approaches to the crucial step of image segmentation are discussed. Cell image segmentation can be seen as a
modeling problem where different approaches are more or less explicitly based on cell models. For example,
thresholding methods can be seen as being based on a model stating that cells have an intensity that is different
from the surroundings. More robust segmentation can be obtained if a combination of features, such as inten-
sity, edge gradients, and cellular shape, is used. The seeded watershed transform is proposed as the most useful
tool for incorporating such features into the cell model. These concepts are illustrated by three real-world prob-
lems.

Pleiades Publishing home page | journal home page | top

If you have any problems with this server, contact webmaster.