v2.0.0

What's Changed

• Enable tracking of detected objects over a multi-year dataset
• wmts tile service, compatible with XYZ connector
• Fix sonarqube code smell in prepare_data and tile-generator by @cleherny in #6
• Adapt the code so to make it compatible with the ac/improve-code-quality branch of the object detector by @acerioni in #8
• Ch/update master od by @cleherny in #10
• Ch/dev tm by @cleherny in #5

New Contributors

• @acerioni made their first contribution in #8
• @GwenaelleSa made their first contribution in #5
• @Shanci-Li made their first contribution in #3

Full Changelog: v1.0.0...v2.0.0

v1.0.0

The project proj-dqry provides a framework to automatically detect Mineral Extraction Sites (MES) in high-resolution images. Images are downloaded with WMS service. Labels are defined and used to train an automatic detection algorithm. The trained model is then used to detect MES in images by inference. The project works along with the object-detector version 1.0.0.

• tile-generator.py defines a custom-size tile fitting an AOI.
• prepare_data.py prepares the data to be compatible with the object-detector scripts.
• prediction-filter.py filters the quarry predictions according to set threshold values.