Open-Source Processing and Analysis of Aerial Imagery Acquired with a Low-Cost Unmanned Aerial System to Support Invasive Plant Management

Frontiers in Environmental Science

Torsten Prinz¹  Jan Thiele²  Tillmann K. Buttschardt²  Jan R. K. Lehmann²  Gustavo Heringer³  João A. A. Meira-Neto³  Silvia R. Ziller⁴ 

[1] Institute for Geoinformatics, University of MuensterMuenster, Germany;Institute of Landscape Ecology, University of MuensterMuenster, Germany;Laboratory of Ecology and Evolution of Plants, Universidade Federal de ViçosaViçosa, Brazil;The Horus Institute for Environmental Conservation and DevelopmentFlorianópolis, Brazil;

关键词: Acacia mangium;    drone;    invasive alien species;    Mussununga;    RPAS;    remote sensing;   

DOI  :  10.3389/fenvs.2017.00044

来源: DOAJ

【 摘 要 】

Remote sensing by Unmanned Aerial Systems (UAS) is a dynamic evolving technology. UAS are particularly useful in environmental monitoring and management because they have the capability to provide data at high temporal and spatial resolutions. Moreover, data acquisition costs are lower than those of conventional methods such as extensive ground sampling, manned airplanes, or satellites. Small fixed-wing UAS in particular offer further potential benefits as they extend the operational coverage of the area under study at lower operator risks and accelerate data deployment times. Taking these aspects into account, UAS might be an effective tool to support management of invasive plant based on early detection and regular monitoring. A straightforward UAS approach to map invasive plant species is presented in this study with the intention of providing ready-to-use field maps essential for action-oriented management. Our UAS utilizes low-cost sensors, free-of-charge software for mission planning and an affordable, commercial aerial platform to reduce operational costs, reducing expenses with personnel while increasing overall efficiency. We illustrate our approach using a real example of invasion by Acacia mangium in a Brazilian Savanna ecosystem. A. mangium was correctly identified with an overall accuracy of 82.7% from the analysis of imagery. This approach provides land management authorities and practitioners with new prospects for environmental restoration in areas where invasive plant species are present.

【 授权许可】

Unknown