Moving Object Detection based on Motion Vector and Color Information Using Omni Camera

Since the omnidirectional camera system with a very large field of view could take many information about environment scene from few images, various researches for calibration and 3D reconstruction using omnidirectional image have been actively reported. The ability to detect moving objects is required for an omni camera to avoid dynamic obstacles and navigate autonomously in real environments with high uncertainty. This paper proposes a method which detects moving objects with a omni camera mounted on the platform. The core of the proposed scheme may be summarized as follows: First, it extracts several candidate regions which can be considered as the moving objects. Second, extracted candidate regions are classified into moving object and background using motion vectors. Third, the omni camera can detect the moving object stably through its color information. It may sometimes be possible to detect the moving objects even in case the motion vectors are not extracted. Experimental results show that the proposed scheme is effective in moving objects detection