Calibration and 3D geometry estimation of a pan-tilt-zoom camera network

Due to the recent growing need for increased surveillance, we believe that it is necessary to move from single or stationary cameras to a network of multiple cameras. This is useful for various applications, like surveillance or 3D model building. In this paper, we deal with such a dynamic network of pan-tilt-zoom (PTZ) cameras. We allow the camera to freely vary its internal parameters by rotating and zooming. Moreover, these cameras need not have an overlapping Field of View (FoV), in order to maximize their area of coverage. A practical framework is proposed to self-calibrate each camera installed in the network. A calibrated camera act as an angle measuring device, allowing for a 3D reconstruction of the scene. It is also shown that only one automatically computed vanishing point and a line lying on any plane orthogonal to the vertical direction is sufficient to infer the dynamic orientation between all the cameras in the network. Using minimal assumptions, we are able to successfully demonstrate promising results on synthetic as well as on real data.