Development of a vision-based feral vertebrate identifier using fuzzy type ii

Abstract

Discrimination between feral vertebrates and livestock is necessary to control the population of the unwanted group of creatures in the large rangelands in Australia. This study proposes an algorithm for feral animal identification in the free environment required prior to a control action to limit the feral animal numbers compete with the livestock on the same natural resources. The inherent uncertainties of an imagery over an arbitrary scene in the nature demands a robust segmentation method. Hence, a set of methods are investigated which turned out that fuzzy logic using fuzzy membership functions is not only more realistic but also produces more reliable animal segmentation.