Predicting On-Road Carbon Emissions Using Traffic Camera Images

This article presents a framework for estimating carbon dioxide (CO2) emissions from highways using images captured at regular intervals by traffic speed cameras, complemented by speed data collected from loop detectors on expressways and major arterial roads. Vehicle counting is performed using the You Only Look Once (YOLO) algorithm to identify and classify cars, buses, and trucks. The results highlight variations in traffic flow and emissions between weekdays and weekends.