This set of scripts and configuration files are related to the quarry/exploitation sites detection case. The detector is initially trained on swissimage from swisstopo using the TLM data of swisstopo for the labels.
For this case, the thermal panel (TPNL) case script is used. See its documentation on the following page
Two files are provided along the prepare script :
- config.yaml example
- logging.conf example
The logging format file can be used as provided. The configuration YAML has to be adapted in terms of input and output location and files.
As the data are prepared using the proposed script, the following procedure can be followed in the appropriate environment :
$ cd [process_directory]
$ python [detector_path]/scripts/generate_training_sets.py [yaml_config]
$ cd [output_directory]
$ tar -cvf images-256.tar COCO_{trn,val,tst}.json && \
tar -rvf images-256.tar {trn,val,tst}-images-256 && \
gzip < images-256.tar > images-256.tar.gz && \
rm images-256.tar
$ cd -
$ python [detector_path]/scripts/train_model.py config_NE.yaml
$ python [detector_path]/scripts/make_prediction.py config_NE.yaml
$ python [detector_path]/scripts/assess_predictions.py config_NE.yaml