Detection of invasive vegetation through UAV and Deep Learning

  • Camargo P. Charles UFSCar
  • Pedro Henrique Corrêa Kim UFSCar
  • Aline Gabriel de Almeida UFSCar
  • Eduardo Vieira Do Nascimentok UFSCar
  • Lidia Gianne Souza Da Rocha UFSCar
  • Kelen Cristiane Teixeira Vivaldini UFSCar


Species originating from one biome are often irregularly introduced into other biomes by accident. This event configures a biological invasion, which can cause irreversible adverse impacts on biodiversity and affect economic productivity in sectors such as fisheries, forestry, and agriculture. Furthermore, many species are vectors of human diseases, making biological invasions a significant problem. In Brazil, monitoring becomes very complex, with many closed forests, such as the mountain regions and other places with difficult accessibility, and demands many resources to maintain it, whether human or financial. Remotely and autonomously detecting invasive vegetation in large or complicated physical access areas can positively impact conservation work. Governments can take concrete actions to favor the environment through this monitoring and avoid irreversible damage to the ecosystem. Therefore, this paper proposes the classification of images using Deep Learning algorithms to detect the invasive species Hedychium Coronarium. We will capture the photos by remote sensing through UAVs (Unmanned Aerial Vehicles).
Palavras-chave: Deep learning, Productivity, Neural networks, Government, Vegetation mapping, Forestry, Unmanned aerial vehicles, Unmanned Aerial Vehicles, Deep Learning, Artificial Neural Networks (ANNs), U-Net
Como Citar

Selecione um Formato
CHARLES, Camargo P.; KIM, Pedro Henrique Corrêa; ALMEIDA, Aline Gabriel de; NASCIMENTOK, Eduardo Vieira Do; ROCHA, Lidia Gianne Souza Da; VIVALDINI, Kelen Cristiane Teixeira. Detection of invasive vegetation through UAV and Deep Learning. In: SIMPÓSIO BRASILEIRO DE ROBÓTICA E SIMPÓSIO LATINO AMERICANO DE ROBÓTICA (SBR/LARS), 13. , 2021, Online. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2021 . p. 114-119.