Files
Abstract
The estimation of wildlife populations is an issue currently being solved at workplaces on many levels. Knowledge of wildlife population and localization is not only very important for reducing damage to agricultural and forest growth, which arises from the local overgrowth of certain animal species, but also for the protection of endangered species of animals and plants. The article presents the results of a research carried out during 2017 as the first partial objective of a complex automated wildlife estimation project, namely the recognition of game in a free landscape without vegetation cover from an UAV (unmanned aerial vehicle). The paper describes a method of finding game animals in a selected area and identifies problems with the recognition of the animals hiding in the vegetation. These results play an important role in solving the overall complex p roblem of automated game recognition.