Wide Area Surveillance with Rotating and Zooming Cameras

Visual surveillance in dynamic scenes is one of the most active research topic in computer vision. Applications include access control, multiple object/person detection, tracking, identity recognition, crowd analysis and anomaly detection. My research has been focused on the problem of real-time, wide area monitoring using rotating and zooming sensors. The algorithms proposed in this work belong to three fundamental classes of problems in computer vision: camera localization and mapping, object detection and object tracking.