Using CNN (convolutional neural networks) to map forest tree species in high resolution UAV-based RGB-imagery.

The portion that will be used for training will be cropped to 256x256 and augmented then used.

Expirement 1: Binary classifier using CNN U-Net

Jump to the notebook here or on GitHub.

DATA :Nguyen, Ha Trang, Diez, Yago, Lopez Caceres, Maximo Larry, Kentsch, Sarah, Moritake, Koma, & Shu, Hase. (2020). Dataset of Individual Sick Fir Detection using Computer Vision and Deep Learning.

Architecture of CNN U-Net: Mapping forest tree species in high resolution UAV-based RGB-imagery by means of convolutional neural networks; Felix Schiefer, Teja Kattenborn, Annett Frick, Julian Frey, Peter Schall, Barbara Koch, Sebastian Schmidtlein; ISPRS Journal of Photogrammetry and Remote Sensing, Volume 170, 2020, Pages 205-215, ISSN 0924-2716;