An active embedded stereo vision system with IR filters for robust indoor localization

In this paper, an active embedded stereo vision system with IR filters is proposed for robust three-dimensional localization. In the system, ordinary cameras with infrared (IR) filters are used instead of expensive IR cameras. IR filtered images could be easily transferred to binary image with low noises. Two control servos do contribution to the pan-tilt movement of the stereo pair. A global search algorithm is then proposed for rapid object localization. Experiments show that the system has a good performance on object localization.