An accident detection system on highway through CCTV with calogero-moser system

Abstract

There are many limits like as shadowing occlusion and no lights in the video image detector systems. In order to make accurate detector system, we need to get rid of these problems specially accident detection system by using vehicle trace. In this paper, we introduce a method of overcoming shadow and nighttime. And we propose the accurate accident detection system. We find the flow of vehicle trace is like as level spacing distribution as Wigner distribution. It is in the level statistics when we represent this vehicle trace avoidable. From this distribution we can derive a probability induced by different of position for each lane. Using this equation we can find abrupt state of vehicle flow comparing with normal flow. We shall show statistical results from some experiments for this system evaluation.