FAW RGB Maize ROS Bag Dataset — Metadata, Manifest, Checksums, and External Raw Data Access

This record provides public documentation, sequence-level metadata, file manifests, and SHA256 checksums for the FAW RGB Maize ROS bag dataset. The dataset corresponds to RGB recordings of maize plants under natural fall armyworm (Spodoptera frugiperda) infestation, acquired in Jerez, Zacatecas, Mexico. The raw ROS .bag files are not stored in Zenodo or GitHub. The complete raw acquisition material consists of five ROS .bag files totaling 8,840,100,688 bytes (8.233 GiB), hosted externally in an institutional read-only OneDrive folder: https://correoipn-my.sharepoint.com/:f:/g/personal/rrevelesm_ipn_mx/IgATA-FW1U91Q5mwSvQGW3kTAQPdW3N31aF5M6g8wnkSDrI?e=kSJ9md The public metadata include file names, sizes, sequence information, and SHA256 checksums to support verification of the downloaded raw files. The dataset is associated with the manuscript “RGB-Based Spectral Indices for Exploratory Assessment of Fall Armyworm (Spodoptera frugiperda) Leaf Damage in Maize: A Case Study in Jerez, Mexico”. The associated study uses extracted RGB frames from these recordings for an exploratory, offline assessment of visible foliar damage using RGB-based image-processing and machine-learning methods. This record does not include independent pixel-level damage annotations, direct insect labels, agronomic severity labels, field-ready validation, or pest-management recommendations. The recordings correspond to a single-site, single-date exploratory acquisition scenario and should not be interpreted as a general-purpose operational benchmark. Data acquisition support: Enrique Díaz Landeros assisted with the field acquisition and recording of the ROS bag sequences used in this study.