ADIAC: Automatic diatom identification and classification

• Unité de recherche: PCRD EU
• Numéro de projet: 98.0048
• Titre du projet anglais: ADIAC: Automatic diatom identification and classification

Mots-clé (Anglais)

Phycology; taxonomy; ecology; image processing; pattern recognition; image retrieval

Programme de recherche (Anglais)

EU-programme: 4. Frame Research Programme - 3.2 Marine sciences and technologies

Description succincte (Anglais)

See abstract

Autres indications (Anglais)

Full name of research-institution/enterprise:
Universität Bern
Institut für Informatik und angewandte Mathematik

Partenaires et organisations internationales (Anglais)

Coordinator: University of Algarve (P)

Résumé des résultats (Abstract) (Anglais)

The ADIAC project is a pilot study concerning the application of image processing and pattern recognition tools to the automation of diatom identification by computer processing.
Diatoms are microscopic algae, found in water and wherever there is enough humidity and light for photosynthesis. There are thousands of different species, and their valves have different shapes and ornamentation. Because of their narrow ecological preferences, they can be used as ecological indicators for measuring pH-values and nutrient status in old sediments. They are studied in archaeology to trace ceramic and clothing fragments. In forensic research they are used to determine the cause of death in drowning cases. Diatoms are very good indicators of water quality, and authorities responsible for drinking water are going to use them to monitor the quality.
The identification of diatoms is a very tedious work, because microscope slides must be prepared, scanned, photographed, and the diatoms found must be linked to described taxa. Usually the latter is done by comparing images of diatoms with those in atlantes, which is quite difficult even for experts because of the very subtle variations of the shapes and ornamentation.
The ADIAC project aims at developing algorithms for an automatic identification of diatoms using image information. Instead of taking only the shape of the valve outline into account, several methods will be applied in order to employ also the valve ornamentation in the identification process. Important aspects are the development of methods for automatic slide scanning, auto focusing, the realization of large image databases, and the integration into taxonomic/ecological database systems.
In the framework of this project, the Research Group of Computer Vision and Artificial Intelligence of the University of Bern is responsible for the development of methods to obtain morphological diatom descriptions and to identify diatoms based on methods of pattern recognition.

Références bases de données (Anglais)

Swiss Database: Euro-DB of the
State Secretariat for Education and Research
Hallwylstrasse 4
CH-3003 Berne, Switzerland
Tel. [phone]
Swiss Project-Number: 98.0048