Depth Map based Object Tracking and 3D Positioning for Non-Static Camera

Mikel Labayen, Julen García, Aritz Legarretaetxebarria, and Maider Laka-Iñurrategi


Image processing applications, depth camera, robotic TV set, automatic object tracking, multiple-object 3D positioning


This paper presents a real-time multi-object tracking and locating system based on visible-light camera and a depth map camera signals. The aim is to obtain the real location of the tracking objects in order to move automatically the source cameras to follow them. One of the big challenges of this approach is the fact that the source cameras will be moving as they follow the target. In addition to this, only one visible-light camera and one depth map camera are used for the tracking of the objects. The system combines image processing techniques and depth-based segmentation what allows more accurate boundaries detection and occlusion facing.

Important Links:

Go Back