Alex Goldhoorn

Publication

← Back to Publications

Searching and Tracking People in Urban Environments with Static and Dynamic Obstacles

Alex Goldhoorn, Anaïs Garrell, René Alquézar and Alberto Sanfeliu
Robotics and Autonomous Systems, 2016
PDF | BibTeX | DOI

Abstract

Searching and tracking people in crowded urban areas where they can be occluded by static or dynamic obstacles is an important behavior for social robots which assist humans in urban outdoor environments. In this work, we propose a method that can handle in real-time searching and tracking of people using a Highest Belief Particle Filter Searcher and Tracker. It makes use of a modified Particle Filter (PF), which, in contrast to other methods, can do both searching and tracking of a person under uncertainty, with false negative detections, lack of a person detection, in continuous space and real-time. Moreover, the method uses dynamic obstacles to improve the predicted possible location of the person. Comparisons have been made with our previous method, the Adaptive Highest Belief Continuous Real-time POMCP Follower, in different conditions and with dynamic obstacles. Real-life experiments have been done during two weeks with a mobile service robot in two urban environments of Barcelona with other people walking around.

Experimental Videos

Videos demonstrating the experiments done with the HB-Particle Filter Searcher & Tracker to search and track people. The robot searches and tracks a person recognized using an adapted version of AR Markers.

Video Legend

Each experiment video shows two sections:

Map Elements (Left)

Probability Map Elements

Limitations

The robot detects the person in two phases: 1) by laser detection of legs combined with Multiple Hypothesis Tracking for Multiple Targets (MHT), and 2) by AR tag. The AR tag alone resulted in some false positives, so we only accepted detected tags if there was a person detection by laser close enough. When false positive detections occur, the probability increases for that location temporarily, but the probability map recovers to the correct area as the false positive is typically only detected briefly.

Experiments

Exploration in a Small Environment

Dabo explored the FME environment without the person present, showing the exploration behavior around obstacles. The belief propagated behind obstacles as the robot moved. A false positive detection occurred when a person with a bike entered, but the belief recovered after a few iterations.

Searching and Tracking

Dabo searched for the person in the FME environment with other people present who occluded the target. The algorithm correctly assumed probability of the person being behind obstacles, and the robot went behind them to verify and then tracked the person.

Exploration in a Larger Environment

Robot exploration on the Telecos Square in different experiments, including handling of false positive detections with belief recovery.

Searching and Tracking in a Larger Environment

Two experiments showing the robot searching for the person on Telecos Square behind obstacles, with tracking while partially occluded.

Searching and Tracking in a Larger Environment 2

Several experiments of the robot searching for and tracking the person on the Telecos Square.

Use of Dynamic Obstacles

This simulation shows 30 people (dynamic obstacles) walking around to random locations. The left side shows the method using visible dynamic obstacles to update the belief, accounting for the person potentially hidden behind them. The right side shows the algorithm without using dynamic obstacles. The belief maps demonstrate how the method maintains probability behind dynamic obstacles.

More Information

The full version of the videos can be found on the full page with all videos.

← Back to Publications