VEHICLE DETECTION USING PROBABILITY FUSION MAPS GENERATED BY A MULTI-CAMERA SYSTEM(International Workshop on Advanced Image Technology 2007)

概要

論文の詳細を見る
The monitoring of traffic is an increasingly important area of research given the increasing numbers of vehicles around the world, leading to problems such as pollution and congestion. To address these problems, traffic management systems are useful. In this paper, we describe a multi-camera traffic monitoring system based on the concept of probability fusion maps for vehicle detection. An inverse perspective mapping is not mathematically defined unless the inverse mapped points are constrained to lie on a horizontal plane. Because road surface points satisfy such a constraint, the images captured by the cameras in the system are inverse mapped into a common reference frame. Thus, the background-subtracted, binarized images are registered and merged to create a "probability fusion map" where the intensity represents the probability of a vehicle.
社団法人電子情報通信学会の論文
2007-01-02