Table 1.xls (5.5 kB)

Computer-vision supported hourly aggregated counting of pedestrian crossings using motion-activated short videos and photographic stills compared to manual in-person counting.

Download (5.5 kB)
posted on 26.10.2020, 18:29 by Evan Thomas, Sally Gerster, Lambert Mugabo, Huguens Jean, Tim Oates

Error estimates for variance and bias indicate a lower error using the videos compared to stills.