Loughborough Urban Pedestrian Count Dataset: Combined Wi-Fi Sensor and Video Survey Data

This dataset contains hourly pedestrian count data for a single representative weekday collected across Loughborough, UK, combining two complementary sources. The first source comprises anonymised hourly counts from eight Wi-Fi sensors positioned across the town centre, each detecting Wi-Fi-enabled devices within an approximate 100-metre radius as a proxy for pedestrian presence. The second source comprises pedestrian counts extracted from video footage at 21 strategic locations distributed across diverse urban morphologies, including the historic core, retail centre, residential neighbourhoods, and university campus. Video counts were derived using the YOLOv7 computer vision model, capturing total pedestrian activity in both directions across all applicable pavements. Both sources were collected over the same 24-hour Monday period during university term-time under stable weather conditions and in the absence of major events, ensuring temporal comparability and minimising confounding factors. The combined dataset provides 29 spatially distributed observation points covering a range of urban contexts, from high-footfall commercial streets to low-footfall residential areas. Counts are aggregated to hourly and daily totals and georeferenced to a common coordinate system aligned with the Loughborough town boundary. Full methodological detail, including sensor coverage zones, YOLOv7 extraction parameters, and spatial processing procedures, is provided in the associated manuscript. Temporal coverage: Single weekday (Monday), university term-timeSpatial coverage: Loughborough, North Leicestershire, UKNumber of observation points: 29 (8 Wi-Fi sensors, 21 video locations)Temporal resolution: Hourly and daily aggregatesFormat: CSV / GeoJSON