Real time aerial natural image interpretation for autonomous ranger drone navigation

In this paper a method for real time interpretation of environmental features for use in visual navigation system of an Unmanned Aerial Vehicle (UAV) is presented. The proposed approach uses manually sampled colour values to classify different land covers. Colour information of a set of manually selected windows is compared to select the best attributes needed for discrimination between different land covers in various (natural) lighting conditions. Each frame is then partially scanned and desired environmental features are extracted and classified. The results show that the proposed technique meets the minimum speed and accuracy requirement of aforementioned application.