A dataset of blowing snow event visibility classification for the Makitasi Wind Zone, China, 2024–2025

Fine-grained visibility grading data are of great significance for model training, control strategy formulation, and the deployment of roadside edge computing devices. Based on continuous videos from four roadside fixed monitoring cameras and visibility instrument observations along the G3015 Keketa Expressway (K264+700–K272+700) in the Makatag Wind Area, Xinjiang, from November 2024 to March 2025, this dataset was constructed by filtering video time based on visibility instrument data, extracting frames at a frequency of 1 minute per frame, and applying manual grading annotation.

The dataset consists of one compressed archive containing seven folders corresponding to seven visibility levels (0–100 m, 100–200 m, 200–300 m, 300–400 m, 400–500 m, 500–1000 m, and 1000–2000 m), comprising 13,124 roadside camera images with a spatial resolution of 640×480 pixels. It also includes one visibility instrument data file in tabular format, containing time, wind speed, wind direction, minimum visibility, and average visibility.