USE OF UAV-BASED RGB IMAGERY IN AGRO-ECOSYSTEMS RESEARCH: A FIRST APPROACH TOWARDS LAND USE CLASSIFICATION

摘要

Spatial information in crop production systems is based on a combination of field reconnaissance and aerial photos or satellite images. Traditional aerial platforms such as planes and satellites are often not sufficient for an accurate land use classification because of their low spatial and temporal resolution (Pajares, 2015). Moreover, the cloudiness and weather conditions often make satellite and aerial products unworkable. Recently, photogrammetry based on unmanned aerial vehicles (UAVs) images has become a feasible, manageable and cost-effective solution to gather very high resolution imagery, resulting in detailed geo-information. In addition, the moment and spatial pattern of respective flight missions can be developed on the spot and focusedon the specific agro-ecosystem under study. Furthermore, many missions on the same location can be planned in time, resulting in longitudinal data, which allow dynamical studies of temporal effects on land uses systems. The main objective of this research is the delineation and classification of UAV based imagery to evaluate land use of particular agro-ecosystems Several missions can be organized efficiently for land use evaluation, resulting in a considerable amount of image based information. As a consequence, an automated UAV imagery data analysis workflow for land use evaluation and classification becomes most necessary.