Skip to main content


A Reproducible Supervised Classification System for Tree Canopy and Deforestation Detection Within an Open Source Python Framework Utilizing NAIP Imagery

Key Words: Deforestation, NAIP, Python, scikit-learn, GDAL, open source, classification

Deforestation monitoring is an essential part of maintaining environments as the loss of forested lands leads to increased amounts of CO2 released into the atmosphere while simultaneously eliminating carbon storage. At smaller scales it leads to both increased runoff rates and subsequently increased erosion, especially in areas where no plant reclamation is initiated. Accurately monitoring deforestation to mitigate these effects on a large scale can be a time consuming and difficult process to complete. Furthermore, commercial software dedicated to completing these tasks such as eCognition or Textron Systems Feature Analyst can be expensive with little insight into how their algorithms are truly working as they are closed source. The lack of knowledge into the inner workings of commercial software leads to the consideration of applying any number of existing open source libraries. Previous research has not effectively published a reproducible and inexpensive method for deforestation monitoring as researchers often do not make the framework available with which the research was conducted. The goal of the proposed research is to create a process that can be repeated to effectively monitor deforestation. With the benefits NAIP imagery possess combined with the effectiveness of the machine learning python library scikit-learn, and GDAL a low level geospatial python library, a dedicated remote sensing method can be developed to solve these issues. Presented will be the preliminary classified results of the system as well as an oral presentation detailing the preliminary framework with which results will be created.


This is a metadata-only record.



  • Subject
    • Environmental Spatial Analysis

  • Institution
    • Gainesville

  • Event location
    • VMR 3 Enter Guest PIN 2003

  • Event date
    • 17 April 2020

  • Date submitted

    19 July 2022

  • Additional information
    • Acknowledgements:

      Huidae Cho