wind-turbine-academic-zt 1.1.0

Streamlined academic tool for wind turbine height estimation using 2D ground shadow models and ephemeris.

🌪️ Academic Wind Turbine Height Estimator (ZT)

A streamlined, industrial-grade Python tool for robust wind turbine height inversion from optical remote sensing imagery.

Unlike traditional methods that struggle with relief displacement (satellite side-looking distortion), this tool strictly utilizes a Pure Ground Shadow Model (Base-to-Shadow Tip) extracted via deep learning. By operating entirely on the 2D ground plane, the system naturally immunizes itself against 3D satellite viewing geometric distortions.

🌟 Core Architectural Breakthroughs

• Zero Distortion Inversion: Relies purely on ground-projected shadow distances, eliminating the need for complex satellite elevation/azimuth compensation.
• Triple-Track Data Injection: Adaptive input routing handling standard metadata, manual overrides, or entirely missing metadata.
• Astronomical Ephemeris Fallback: Includes a built-in orbital mechanics engine. Provide the spatial coordinates and time, and the system mathematically resolves the solar geometry—no metadata required.

🛠️ Environment Initialization

pip install --upgrade wind-turbine-academic-zt

(Note: Ensure PyTorch and Ultralytics YOLOv8 are properly configured in your environment before execution).

🚀 Execution Protocols (Triple-Track Architecture)

The system operates via a unified CLI entry point. You must provide the target image (--tiff) and its Ground Sample Distance (--gsd).

Choose ONE of the three operational tracks based on your available data:

Track 1: Automated XML Parsing (Standard)

Automatically extracts solar geometry from standard satellite metadata (e.g., GF-2). Includes anomaly interception for known telemetry issues (e.g., Zenith/Elevation confusion).

python main.py --tiff data/demo.tif --gsd 0.8 --xml data/demo.xml

Track 2: Manual Angle Injection

Bypass metadata extraction entirely and inject a known solar elevation angle directly.

python main.py --tiff data/demo.tif --gsd 0.8 --sun_ele 63.18

Track 3: Astronomical Ephemeris Mode (Metadata-Free)

If metadata is lost or corrupted (e.g., drone imagery, historical archives), provide the geospatial and temporal coordinates. The built-in ephemeris engine will resolve the solar angle mathematically.

python main.py --tiff data/demo.tif --gsd 0.8 --ephemeris --lat 38.25 --lon 107.10 --time "2026-05-25 04:00:00"

📊 Output Artifacts

Upon successful execution, verified artifacts are generated in the output/ directory:

• final_heights.csv: Contains extracted Pixel Distances, Ground Shadow Lengths, and Calculated Physical Heights (m).