Combining Passive Visual Cameras and Active IMU Sensors to Track Cooperative People


We attack the problem of persistently tracking cooperative people such as children, the elderly or patients by combining passive tracking and active tracking techniques. Passive tracking uses visual signals from surveillance cameras, but vision based people tracking becomes a hard problem in challenging scenarios such as long-term/heavy occlusion, people changing their movement patterns during occlusion, or people temporarily moving out of the visual field. Active tracking uses sensor signals from Inertial Measurement Unit (IMU) carried by targets themselves. IMU-based tracking is independent of visual signals, so it keeps working when people are visually occluded and offers clues where the target could be, helping the visual tracking to reidentify the target. Meanwhile, when visual signals on people are available, visual tracking can calibrate IMU-based tracking to avoid sensor drift. The experimental results show that the IMU and visual tracking are complementary to each other and their combination performs robustly on tracking cooperative people in many challenging scenarios.

Meeting Name

2015 18th International Conference on Information Fusion (2015: Jul. 6-9, Washington, DC)


Computer Science

Keywords and Phrases

Cameras; Information Fusion; Security Systems; Tracking (Position); Units of Measurement; Active Tracking; Inertial Measurement Unit; Movement Pattern; Passive Tracking; People Tracking; Sensor Signals; Surveillance Cameras; Visual Tracking; Target Tracking

International Standard Book Number (ISBN)


Document Type

Article - Conference proceedings

Document Version


File Type





© 2015 Institute of Electrical and Electronics Engineers (IEEE), All rights reserved.