Releases: Nasef2017/Bathymetrix-AI
Bathymetrix_AI V5.0
1️⃣ Advanced 3-Indices Water Masking: We've implemented a robust, fully automated water masking algorithm that leverages 3 indices (NDWI, MNDWI, and NWI) alongside adaptive thresholding. This ensures the perfect isolation of the aquatic domain, drastically reducing false positives from land and clouds.
2️⃣ Customizable Deep Water Filter: We've introduced a brand-new deep water filter that automatically calculates statistical thresholds to isolate Optically Shallow Water (OSW), ensuring we only model depths where the satellite signal is truly valid.
3️⃣ Full ML Algorithm Customization: The deep water filter and the entire machine learning pipeline are now fully customizable! You have complete control over the hyperparameters for all our built-in algorithms (Random Forest, Gradient Boosting, SVR, Extra Trees, etc.), empowering you to fine-tune the Auto-ML models precisely for your unique coastal region.
Bathymetrix_AI V4.8
Fix Reporting Bug
Bathymetrix_AI V4.7
Enhance Stability. Added full compatibility for QGIS 4.0 (Qt6/PyQt6) via qgis.PyQt
Bathymetrix_AI V4.4
Enhance Stability
Bathymetrix_AI V4.3
Enhance Python code Stability
v4.2
Bathymetrix-AI v4.1
Advanced Data Filtering: Integrated Density-based and Statistical Outlier Removal (SOR) alongside RANSAC for superior ICESat-2 seabed photon isolation.
Customizable Feature Stacking: Enabled flexible selection of spectral and spatial input features prior to Phase 2, offering full control over the model's training variables.
Bathymetrix_AI V4.0
Add Hyperparameter Optimizers ( Random Search, Gird Search, Gaussian).
Add ways to handle exiting 2 points or more into the same pixel.
Make specific Algorithms Hyperparameter range editable from the SDB Master_Flow toolbox.
Enhance Provider.py to be more stable.
Bathymetrix_AI V3.3
Update Help, Readme for easier usage
Bathymetrix_AI V3.2
Enhancement for Phase 04 and Phase 02