Application of modified filtering algorithm to signalized urban arterial street incident detection.

A few studies have reported on algorithms to detect incidents on arterial streets. These reported algorithms are based on new technology, such as video-image processing, and require closely spaced detector stations in a link. In this paper, a real-time Kalman filtering algorithm was adopted based on the findings of existing freeway and arterial street incident detection algorithms. A field test and TRAF-NETSIM simulation were performed to verify the algorithm.