Department of Image Processing

The aim of stereology is to estimate 3-dimensional geometric properties of particles based on lower-dimensional probes, e.g. to estimate the particle volume or more complicated characteristics based on planar sections. One of the surprising and not-so-intuitive results of local stereology is an unbiased estimator of particle volume from a single section or even a single line section through the particle, that is a method of volume estimation that gives "on average" the correct result using only very little information. Similarly, it is possible to estimate the particle surface area in an unbiased way based on a single planar section. Apart from a simple introduction to local stereology we will also present a semi-automatic method of surface area estimation taking advantage of automatic segmentation of digital images. We illustrate the method on real data - images of neurons from mice hipocampi.