Multi Object Detection and Tracking from Video File

Rapee Krerngkamjornkit, Milan Simic
  • Applied Mechanics and Materials, February 2014, Trans Tech Publications
  • DOI: 10.4028/www.scientific.net/amm.533.218

Detection and Tracking of Objects From Video File Source

What is it about?

This paper describes computer vision algorithms for detection, identification, and tracking of moving objects in a video file. The problem of multiple object tracking can be divided into two parts; detecting moving objects in each frame and associating the detections corresponding to the same object over time. New Video tracking algorithm was successfully tested using the standard walking pedestrians datasets.

Why is it important?

The experimental results show that system can operate in real time and successfully detect, track and identify multiple targets in the presence of partial occlusion.

Perspectives

Dr Milan MS Simic (Author)
RMIT University, School of Engineering

The likelihood of tracking errors can be reduced by using a more complex motion model, such as constant acceleration, by using an Extended Kalman filter or Partical filter. Also, other cues could be incorporated for associating detentions over time, such as size, shape, and colour.

The following have contributed to this page: Dr Milan MS Simic