Real-Time People Tracking and Counting Based on Computer Vision

This project addresses the development of a vision-based system for real-time people counting in indoor and outdoor scenes. Unlike most existing methods, which are frequently based on the use of sensor sets, such as photoelectric and pressure sensors, this work proposes a new technique based on the application of an intelligent system for analyzing images of the monitored scene, captured by a single steady camera (possibly embedded in mobile devices, such as smartphones). Therefore, our solution does not need to be used in conjunction with physical structures, which may constrain the space where people flow and force them to move through specified paths.

---

Publications

NEURO 2013

SBAI 2009

Master Thesis – Silva, L.

Term Paper – Pereira, R.

---

See also