Abstract-We present a visual sensor network, comprising wide field-of-view (FOV) passive cameras and pan/tilt/zoom (PTZ) active cameras, which automatically captures high quality surveillance video of selected pedestrians during their prolonged presence in an area of interest. A wide-FOV static camera can track multiple pedestrians, while any PTZ active camera can follow a single pedestrian at a time. The proactive control of multiple PTZ cameras is required to record seamless, high quality video of a roaming individual despite the observational constraints of the different cameras. We formulate PTZ camera assignment and handoff as a planning problem whose solution achieves optimal camera assignment with respect to predefined observational goals.