Fuzzy clustering in avian infrared imagery application

Autor: Golrokh Mirzaei, Mohsin M. Jamali, Verner P. Bingman, Peter V. Gorsevski, Jeremy D. Ross
Rok vydání: 2014
Předmět:
Zdroj: EIT
DOI: 10.1109/eit.2014.6871768
Popis: There are large number of reports regarding bird and bat mortality due to strikes with turbine blades in wind farm applications. This issue is threatening the avian life especially migratory birds and bats. Avian monitoring techniques can be used to detect bird and bats, assess their activity, and make intelligent decision for construction of wind farms. In this paper, an IR monitoring approach is used for monitoring birds and bats in an area potential for future construction of wind turbines. As there is no a priori database of images for local birds/bats, clustering is used as an effective technique to group different avian categories. Fuzzy C Means is developed in this application to cluster the targets into birds, bats, and insects groups. The results are then compared with a hard Ant Clustering Algorithm. The number of created clusters and number of individuals in each cluster are then compared.
Databáze: OpenAIRE