Skip to main content

Zeyu Fu

Improving data association in online human tracking.

Email: z.fu2@ncl.ac.uk

Project title

Novel data association methods for online multiple human tracking

Supervisors

Project description

Video-based multiple human tracking plays a crucial role in many applications such as:

  • intelligent video surveillance
  • human behaviour analysis
  • health-care systems

The detection based tracking framework has become the dominant paradigm in this research field. The major task is accurate data association between detections across the frames.

Online multiple human tracking relies on detection in real time for data association. It becomes more challenging with noisy detections, missed detections, and occlusions.

We will address existing challenges in human tracking. We will explore various signal processing, machine learning and information fusion methods.

Interests

Machine learning, video processing, information fusion and medical image analysis.

Publications

Qualifications

  • BEng in Electrical and Electronics Engineering (1st class), Newcastle University, UK.