Identifying High-accuracy Regions in Traffic Camera Images to Enhance the Estimation of Road Traffic Metrics: A Quadtree-based Method

The growing number of real-time camera feeds in urban areas has made it possible to provide high-quality traffic data for effective transportation planning, operations, and management. However, deriving reliable traffic metrics from these camera feeds has been a challenge because of the limitations of current vehicle detection techniques, as well as the various camera conditions, such as height and resolution. In this work, a quadtree-based algorithm is developed to continuously partition the image extent until only regions with high detection accuracy remain. These regions are referred to as high-accuracy identification regions (HAIRs) in this paper. We demonstrate how the use of HAIRs can improve the accuracy of traffic density estimates using images from traffic cameras at different heights and resolutions in Central Ohio. Our experiments show that the proposed algorithm can be used to derive robust HAIRs in which vehicle detection accuracy is 41% higher than that in the original image extent. The use of HAIRs also significantly improves the traffic density estimation with an overall decrease of 49% in root mean squared error.