Title :
Tracking people within groups with RGB-D data
Author :
Munaro, Matteo ; Basso, Filippo ; Menegatti, Emanuele
Author_Institution :
Dept. of Inf. Eng., Univ. of Padova, Padua, Italy
Abstract :
This paper proposes a very fast and robust multi-people tracking algorithm suitable for mobile platforms equipped with a RGB-D sensor. Our approach features a novel depth-based sub-clustering method explicitly designed for detecting people within groups or near the background and a three-term joint likelihood for limiting drifts and ID switches. Moreover, an online learned appearance classifier is proposed, that robustly specializes on a track while using the other detections as negative examples. Tests have been performed with data acquired from a mobile robot in indoor environments and on a publicly available dataset acquired with three RGB-D sensors and results have been evaluated with the CLEAR MOT metrics. Our method reaches near state of the art performance and very high frame rates in our distributed ROS-based CPU implementation.
Keywords :
image classification; image sensors; mobile robots; multiprocessing systems; object detection; robot vision; target tracking; CLEAR MOT metrics; ID switches; RGB-D data; RGB-D sensor; depth-based subclustering method; distributed ROS-based CPU implementation; high frame rates; indoor environments; limiting drifts; mobile platforms; mobile robot; online learned appearance classifier; people detection; robust multipeople tracking algorithm; three-term joint likelihood; Cameras; Detectors; Image color analysis; Robot sensing systems; Target tracking;
Conference_Titel :
Intelligent Robots and Systems (IROS), 2012 IEEE/RSJ International Conference on
Conference_Location :
Vilamoura
Print_ISBN :
978-1-4673-1737-5
DOI :
10.1109/IROS.2012.6385772