Crowd Motion Analysis from Video

People: Gayathri Mahalingam, Chandra Kambhamettu and Benigno Aguirre

Crowd analysis finds its usage in large number of applications like crowd management, public space design, virtual environments, visual surveillance and intelligent environments. The Computer Vision methods focus on extracting quantitative features and in object detection and tracking. Occlusion caused by the high clutter of people in the crowd scene is the major challenge for crowd detection and tracking. The various light intensities in the video and camera motions also make the tracking a challenging problem in the field of computer vision. We use a simplified object tracking method to effectively track people in a crowd scene. The technique robustly track people under various lighting conditions and occlusions.

Related publications

  • Gayathri Mahalingam, Chandra Kambhamettu, Benigno Aguirre, Crowd Motion Analysis from Video. Proceedings of NSF Research Engineering and Innovation Conference, 2009.