Tracking cycle: a new concept for simultaneous tracking of multiple moving objects in a typical traffic scene

Abstract

Motion detection from a moving observer has been a very important technique for 3D dynamical image analysis, especially in the research of obstacle detec- tion and tracking for autonomous driving systems and driver supporting systems. Because of the complicated motion pattern of moving objects in a typical on-road traffic scene, detecting the real moving objects and properly tracking each one have become very difficult and always employs optical flows and fixed template matching to detect and track the moving objects. Un- fortunately, it has to pay a huge calculation cost and can not deal with some complicated motion pattern like overlapping of tracked objects. A new concept of tracking cycle is proposed in this paper. With the con- tinuously estimation of moving objects Vitality and reliability values, our method can detect and track up to more than 30 moving objects in real time. The flexible dynamic template matching method is also presented in this paper, which makes our system faster and more efficient for real-time ITS applications. Ex- periments on real outdoor road scene show the accu- racy and efficiency of our approach.