1A2-B17 Development of Vegetation Survey System using Visible / Near-infrared Images Gathered by a Small Unmanned Aerial Vehicle
Abstract
This paper describes a vegetation survey system using a small Unmanned Aerial Vehicle (UAV). A handy and effective means to observe vegetation for the agriculture and forestry environment are highly desirable. Accordingly, a more useful and low-cost vegetation survey system than traditional methods which use satellites or manned air planes is developed by our research. This system autonomously collects the aerial images using the visible and near-infrared digital cameras equipped on the UAV. In addition, this system makes a widespread mosaic image of aerial pictures, and makes land cover classification image of a mosaic image by the unique classification method based on the k-means method. The field experiments were conducted at Kirigatani moor in Hiroshima Prefecture. It was confirmed that this system is effective and low-cost for the vegetation survey by the results of experiments.