Phytoplankton identification by combined methods of morphological processing and fluorescence imaging M. Lauffer, F. Genty, J.L. Collette CentraleSupelec LMOPS, IMS Metz, France mathieu.lauffer@centralesupelec.fr S. Margueron Université de Lorraine LMOPS Metz, France Abstract— The identification of phytoplankton is currently an important issue to prevent the aquatic environment. The growth of one or several phytoplankton species can lead to hyper eutrophication and causes lethal consequences on other organisms. In this paper, the selective recognition of invading species is investigated by automatic recognition algorithms of optical and fluorescence imaging. Firstly, morphological characteristics of algae of microscopic imaging are treated. The image processing lead to the identification the genus of aquatic organisms and compared to a morphologic data base. Secondly, fluorescence images allow an automatic recognition based on multispectral data that identify locally the ratio of different photosynthetic pigments and gives a unique finger print of algae. It is shown that the combination of both methods are useful in the recognition of aquatic organisms. Keywords— Image processing, fluorescence, phytoplankton recognition Introduction (Heading 1) Phytoplankton and particularly algae plays a fundamental role in the living world. It is a dioxygen generator and the most important carbon dioxide fixer on Earth. However, under certain conditions, their development may become so excessive that it could be harmful to others vegetals and animals of aquatic life, this phenomenon is called the hyper eutrophication [1,2]. Such a situation leads to dramatic consequences on environment due to the difficulties of other plant species to continue photosynthesis and gas exchanges. This can provoke the death of the whole aquatic ecosystem. Moreover, many of these invasive species can present also a risk for human health. Especially, many kinds of cyanobacteria are potentially toxic due to their ability to produce toxins. Both animal (fish, cattle) and human poisonings have been unfortunately observed [3]. It appears therefore essential to strengthen the vigilance on controlling the proliferation of plankton and toxins with the necessity of risk evaluation. In a first approach, the recognition and the identification of aquatic organisms are performed by specialists algologists from microscopic observations. Nevertheless, in certain circumstances, it may be useful to dispose of an automatic recognition system to improve the monitoring of high-risk water ponds and to optimize human intervention of specialists algologists. The development of such an automatic system of recognition of aquatics organism is more and more considered [4]. In order to identify aquatic organisms, an optical microscope in transmission was developed in which the incident and emitted light beams can be filtered in wavelengths. The set up enables the acquisition of microscopic images and multispectral images of fluorescence emission under different illumination. These different images are then analyzed and processed by two algorithms to collect morphologic and pigments characteristics. The first algorithm consists in extracting the shape of the algae. The second algorithm consists in a point-by-point analysis of the fluorescence intensities. Finally, the characteristic parameters linked to morphology and fluorescence emission of the algae are collected to build a database useful for automatic optical recognition. I. MATERIALS AND METHODS A. Fluorescence imaging set up The fluorescence imaging system is a home-made set-up operating in transmission with inverted optics and illumination by the top (Fig.1). The excitation is a ultra-violet/visible lamp (OmniCure S1000) which presents a average irradiance of 50 mW/cm² on the sample. The excitation can be filtered by several band-pass filters of 10 nm at 400 nm, 440 nm and 470 nm [5-7]. The samples are analyzed on sapphire cover to prevent ultraviolet absorption of glass. The acquisitions are performed with x20 microscope objective. The fluorescence emission is analyzed from 650 nm to 700 nm. The collected emission between 650 and 700 nm is done according to chlorophylls emission [8-10]. Fig. 1. Fluorescence imaging system in transmission developed for the imaging of phytoplankton BioCapTech Project, Région Lorraine.