COUNTING EUCALYPTUS TREES USING UAV AND DIGITAL IMAGE PROCESSING

Autores

  • Arthur Avelar Chaves Instituto de Ciências Matemáticas e de Computação - Universidade de São Paulo (ICMC/USP)
  • Rodrigo Antonio La Scalea Instituto Federal de Educação, Ciência e Tecnologia de São Paulo - IFSP/São Carlos
  • Célia Leiko Ogawa Kawabata Instituto Federal de Educação, Ciência e Tecnologia de São Paulo - IFSP/São Carlos
  • Kalinka Regina Lucas Jacquie Castelo Branco Instituto de Ciências Matemáticas e de Computação - Universidade de São Paulo (ICMC/USP)

Resumo

New technologies are being used in precision agriculture, like the use of UAV for acquiring images with high spatial and temporal resolution, which allow for a better monitoring of thecrop. In this study, we aim to identify and quantify Eucalyptus trees in aerial images obtained with a UAV. To do so, we implemented digital image processing algorithms with the OpenCV library. The proposed method successfully identified 342 trees in an area of 2025 m², showing to be satisfactory, since that it did not miss any well­defined tree and the estimative for dense groups of trees is fairly accurate.

Downloads

Publicado

31.12.2018