A Radiometrically-Traceable Pipeline for UAV Thermal Imagery with Dynamic Atmospheric Correction

Le résumé fourni par la source

This framework provides a unified GUI and CLI-based workflow designed to convert UAV R-JPEG files into 32-bit floating-point (float32), radiometrically calibrated, and GIS-ready TIFF rasters. By integrating dynamic atmospheric correction (via NASA POWER and Copernicus ERA5) and leveraging the DJI Thermal SDK, the tool preserves the native 14-bit sensor resolution, effectively mitigating quantization artifacts. The modular Python architecture ensures high-throughput operational efficiency for geoscientific research, bridging the gap between raw physical measurements and high-precision geospatial models.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.