Water monitoring: automated and real time identification and classification of algae using digital microscopy

Abstract

Microalgae are unicellular photoautotrophs that grow in any habitat from fresh and saline water bodies, to hot springs and ice. Microalgae can be used as indicators to monitor water ecosystem conditions. These organisms react quickly and predictably to a broad range of environmental stressors, thus providing early signals of a changing environment. When grown extensively, microalgae may produce harmful effects on marine or freshwater ecology and fishery resources. Rapid and accurate recognition and classification of microalgae is one of the most important issues in water resource management. In this paper, a methodology for automatic and real time identification and enumeration of microalgae by means of image analysis is presented. The methodology is based on segmentation, shape feature extraction, pigment signature determination and neural network grouping; it attained 98.6% accuracy from a set of 53 869 images of 23 different microalgae representing the major algal phyla. In our opinion this methodology partly overcomes the lack of automated identification systems and is on the forefront of developing a computer-based image processing technique to automatically detect, recognize, identify and enumerate microalgae genera and species from all the divisions. This methodology could be useful for an appropriate and effective water resource management.