Automatical Analysis of Water Specimens for Phytoplankton Structure Estimation ?

An automatic microscope scanning and recognition system on base of the Utermöhl method is developed for image acquisition, archiving (optical fixation) and phytoplankton analysis. The system, called PLASA (PLAnkton Structure Analysis), enables a characterization of water specimens by their populations. It is described in detail with focus on the image analytical aspects. Plankton chambers are scanned by selectable grid and objective(s). Acquisition positions are automatically focussed and digitized by a TV camera in bright field and, using samples adequately fixed (e.g. with glutaraldehyde), in fluorescence. Interactive programs for design of training sets, image analysis with numerous quantitative features and automatical classifications for a number of organisms are developed and implemented. A long term experiment (23 weeks, 4 locations for water specimens) is performed to generate a reliable data set for training and testing purposes. These data are used to present exemplarily some results for phytoplankton structure characterization. PLASA presents an automated system, comprising all steps from sample processing to algae identification up to species level and quantification.