In cell biology confocal laser scanning microscopic images of the actin filament of human osteoblasts are produced to assess the cell development. This thesis aims at an advanced approach for accurate quantitative measurements about the morphology of the bright-ridge set of these microscopic images and thus about the actin filament. Therefore automatic preprocessing, tagging and quantification interplay to approximate the capabilities of the human observer to intuitively recognize the filaments correctly. Numerical experiments with random models confirm the accuracy of this approach.
