How to count and size fluorescent microbial plankton with digital image filtering and segmentation
Abstract
A fast, reproducible, completely automatic and rather accurate way of quantitative determination of fluorescent microbial plankton is shown. In addition to the question about the organization of an image analyser, which should nearly work without any human intervention, some essential effects on enumeration and sizing of fluorescent cells are treated. Fine tuning between adjustment of the optical environment (camera intensity, dye concentration, focus level, magnification) and image interpretation algorithms as well as an object adapted application of image segmentation techniques (edge detection) are necessary. The edge detector should be able to separate adjacent objects on the basis of significant grey level distributions and changes within local image areas. Sources of error mainly caused by limited image resolution are made more transparent, whereby it is easier for the user to judge results of image analysis in epifluorescence microscopy.