The Autonomous Image Analyzer - enumeration, measurement and identification of marine phytoplankton

An inexpensive, automated system is described for counting, measuring and identifying objects in an aqueous suspension. The results of counting and measurement obtained from the image analyzer were compared to results from visual analysis and a sensor analyzer (HIAC/ROYCO). For a mixture of 3 algal species, each having a characteristic size and shape (Prorocentrum micans, Nitzchia closterium and Hymenomonas elongata), the sensor analyzer failed to distinguish among the 3 species, while results obtained by the image analyzer and with visual methods did not differ significantly; the analyzer identified and counted each of them accurately. If the resolution of the input image is adequate, the image analyzer can identify and classify up to 26 different forms in one sample. The operator need not be present while the samples (up to 40) are being processed. The discussion considers the extension of the method for the counting and sizing of objects with relatively complex geometries like copepods and other zooplankton.