====== Open source canopy classification ======
[[open_source_canopy_classification:start|Workspace]]

{{ :down_atlanta.png?nolink&300|Downtown Atlanta, GA classified}}

This project is parallel research to the [[gfc_canopy_assessment|GFC canopy assessment]] by 
[[owen_smith|Owen Smith]] for [[huidae_cho|Dr. Cho's]] [[https://hcho.isnew.info/gisc-4903.html|GISC 4903 - Special Topics in GIS]]  .
Its focus is on creating an open source canopy classification system designed 
specifically to utilize [[https://www.fsa.usda.gov/programs-and-services/aerial-photography/imagery-programs/naip-imagery/|NAIP imagery]]
within an efficient Python module built on 
[[https://scikit-learn.org/stable/|Scikit-learn]], 
[[https://gdal.org/|GDAL]], and 
[[https://numpy.org/|NumPy]].

The goal is to create an alternative to proprietary classification systems (e.g., Feature Analyst by Textron Systems) to be used for ongoing research into canopy change and the need for efficient data creation that is required.

===== Presented in the 2020 UNG Annual Research Conference =====

  * [[https://youtu.be/6b1n374iBwI?t=4792|Talk]]
  * [[https://ungprod-my.sharepoint.com/:p:/g/personal/ocsmit7654_ung_edu/EZt3OuJGmkxCsf4zT5MAGesBX_ovLA6UZRVKAw0F5ZGEjQ?e=V72Aat|Slides]]

===== Publications =====

  * {{ ::canopy_foss_poster.pdf | Poster}}
  * {{:a_random_forests_based_canopy_classification_system_utilizing_naip_imagery_within_a_python_framework.pdf | Paper}} 

===== Software =====

The source code for the project can be viewed in the
[[https://github.com/ocsmit/canoclass|CanoClass GitHub repository]].

{{tag>projects}}