Tracking and measuring a moving object with a binocular camera system

In this paper, new approaches to track and measure a moving object in 3D space are presented with a binocular camera system. As the moving object out of the field of view, one or two cameras of the pan-tilt binocular system are controlled to search the object, and compute its position in a world coordinate system. Tracking and measuring of a specified object can be done in real time as the cameras moving, without recalibrating the two cameras. Adaptive background model and CamShift algorithm are used respectively to identify moving objects. Experimental results are shown here to demonstrate the effectiveness of the proposed methods.