CamelinaWeed: Annotated UAV Imagery Dataset for Camelina sativa

This dataset provides UAV-based RGB and multispectral imagery for crop monitoring, weed mapping, and field-level analysis in Camelina sativa cultivation. Data were collected from agricultural fields in Thessaloniki and Chalkidiki, Greece, during summer 2025 and winter 2025–2026. The dataset includes manually annotated RGB UAV images with expert agronomist-generated polygon annotations of weed instances, raw RGB and multispectral UAV imagery, and RGB and multispectral orthomosaic products in GeoTIFF format. The annotations include broadleaf, narrowleaf, and species-level weed labels, supporting weed detection, classification, semantic segmentation, instance segmentation, hierarchical learning, and weed distribution analysis. The data were acquired using DJI Phantom 4 Pro and DJI Mavic 3M UAV platforms at different flight altitudes and spatial resolutions. The dataset is intended to support computer vision and precision agriculture research under realistic field conditions. Dataset Download and Reconstruction Instructions Due to the large size of the dataset, the archive is distributed in multiple split parts: CamelinaWeed.zip.part-aa CamelinaWeed.zip.part-ab CamelinaWeed.zip.part-ac CamelinaWeed.zip.part-ad CamelinaWeed.zip.part-ae Please download all parts before reconstruction. Linux / macOS Reconstruct the original archive using: cat CamelinaWeed.zip.part-* > CamelinaWeed.zip Then extract the dataset with: unzip CamelinaWeed.zip Windows Download all archive parts into the same folder and open the first file (CamelinaWeed.zip.part-aa) using 7-Zip: https://www.7-zip.org/ 7-Zip will automatically reconstruct and extract the complete archive.