Skip to content

Latest commit

 

History

History
53 lines (37 loc) · 2.19 KB

bird_detector.md

File metadata and controls

53 lines (37 loc) · 2.19 KB

Bird Detector

Utilizing the same workflow as the tree detection model, we have trained a bird detection model for airborne imagery.

m = main.deepforest()
m.use_bird_release()

We have created a GPU colab tutorial to demonstrate the workflow for using the bird model.

For more information, or specific questions about the bird detection, please create issues on the BirdDetector repo

Annotating new images

If you would like to train a model, here is a quick video on a simple way to annotate images.

<iframe src="https://www.loom.com/embed/e1639d36b6ef4118a31b7b892344ba83" frameborder="0" webkitallowfullscreen mozallowfullscreen allowfullscreen style="position: absolute; top: 0; left: 0; width: 100%; height: 100%;"></iframe>

Using a shapefile we could turn it into a dataframe of bounding box annotations by converting the points into boxes

df = shapefile_to_annotations(
    shapefile="annotations.shp", 
    rgb="image_path", box_points=True, buffer_size=0.15
)

Optionally we can split these annotations into crops if the image is large and will not fit into memory. This is often the case.

df.to_csv("full_annotations.csv",index=False)
annotations = preprocess.split_raster(
    path_to_raster=image_path,
    annotations_file="full_annotations.csv",
    patch_size=450,
    patch_overlap=0,
    base_dir=directory_to_save_crops,
    allow_empty=False
)

Citation

A general deep learning model for bird detection in high resolution airborne imagery Ben G. Weinstein, Lindsey Garner, Vienna R. Saccomanno, Ashley Steinkraus, Andrew Ortega, Kristen Brush, Glenda Yenni, Ann E. McKellar, Rowan Converse, Christopher D. Lippitt, Alex Wegmann, Nick D. Holmes, Alice J. Edney, Tom Hart, Mark J. Jessopp, Rohan H Clarke, Dominik Marchowski, Henry Senyondo, Ryan Dotson, Ethan P. White, Peter Frederick, S.K. Morgan Ernest bioRxiv 2021.08.05.455311; doi: https://doi.org/10.1101/2021.08.05.455311

https://www.biorxiv.org/content/10.1101/2021.08.05.455311v1.abstract