JCURA 2024 PyLC Image Testing Set and Landcover Masks

This is the image and landcover mask testing set created for the 2024 JCURA project "Mountains of Confusion: Evaluating Image Enhancement to Improve AI Landscape Classification" by Larissa Bron. 

The Python Landscape Classification tool (PyLC) [https://github.com/scrose/pylc] was trained using 95 images from two researcher's work, [Fortin (2018)](https://dspace.library.uvic.ca/items/0a911eb0-53bf-4a82-a75a-8b6949c28edd) and [Jean et al. (2015)](https://ieeexplore.ieee.org/document/7045940), and tested with 19 images that were a combination of 11 images from Fortin and Jean withheld from training and 8 images from the Landscapes in Motion project [(Higgs et al., (2020))](https://friresearch.ca/publications/advances-visual-applications-visualizing-quantifying-landscape-change-sw-alberta-using). This core set of training and testing images was used to train the colour models of PyLC with repeat images, then these images were grayscaled and added to training the grayscale models of PyLC with historic images. 

For this JCURA project, the testing image set was updated to: 1) Incorporate images from more geographic areas, and 2) Remove images with reference land cover masks with large errors. 

The files included are 24 test images (.jpg, .tiff, .tif) and 24 manually annotated land cover masks (.png). File names indicate: Researcher_ImageIdentifier_Time where time is whether the image is a historic capture or a repeat.