High-precision traffic congestion detection using lane-by-lane polygons

Abstract

Traffic congestion near shopping centers or traffic accidents may be avoided by keeping away from specific lanes. Therefore, this study proposes a large-scale distributed data processing technique and a polygonal-shape based data search technique to detect traffic congestion at a higher accuracy than VICS. The implementation and experiments suggest establishment of a semi-real-time lane-by-lane traffic congestion detection method possibly be used for nationwide vehicle data.