Multi-robot multiple hypothesis tracking for pedestrian tracking

Nicolas A. Tsokas, Kostas J. Kyriakopoulos

Research output: Contribution to journalArticlepeer-review


In this paper the problem of tracking walking people with multiple moving robots is tackled. For this purpose we present an adaptation to the Multiple Hypothesis Tracking method, which unlike classic MHT, allows for one-to-many associations between targets and measurements in each hypothesis production cycle and is thus capable of operating in a scenario involving multiple sensors. Derivation of hypotheses probabilities accounts for the continuously changing overlapping areas in fields of view of the robots sensors and for detection uncertainty. In the context of three experiments involving people walking among moving robots, the successful integration of our tracking algorithm to a real-world setup is assessed.

Original languageEnglish (US)
Pages (from-to)63-79
Number of pages17
JournalAutonomous Robots
Issue number1
StatePublished - Jan 2012


  • Laser-based range sensing
  • Multi-robot target tracking
  • Multiple hypothesis tracking
  • Pedestrian tracking

ASJC Scopus subject areas

  • Artificial Intelligence


Dive into the research topics of 'Multi-robot multiple hypothesis tracking for pedestrian tracking'. Together they form a unique fingerprint.

Cite this