The purpose of this tool is to create sky masks to improve photogrammetric reconstruction.
git clone https://github.com/OpenDroneMap/SkyRemoval.git
cd SkyRemoval
pip install -r requirements.txt
Usage:
python skyremoval.py source dest
It will automatically download the pre-trained model and run the processing.
source can be a folder or a single image file. dest should be a folder.
usage: skyremoval.py [-h] [--model MODEL] [--ignore_cache] [--in_size_w IN_SIZE_W] [--in_size_h IN_SIZE_H] source dest
SkyRemoval
positional arguments:
source Source image path, can be a single image or a folder
dest Destination folder path
optional arguments:
-h, --help show this help message and exit
--model MODEL Model path, can be a URL or a file path
--ignore_cache Ignore cache when downloading model
--in_size_w IN_SIZE_W
Trained model input width
--in_size_h IN_SIZE_H
Trained Model input height
If you want to speed up the processing, you can use CUDA. Check out the following link for more information: https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html