Automatic video segmentation and object tracking with real-time RGB-D data
Journal
IEEE International Conference on Consumer Electronics
Pages
486-487
Date Issued
2014
Author(s)
Abstract
In this paper, we propose a novel system that is able to automatically segment objects in real-time video by RGB-D information without human annotation (labeling). And it can track each cluster as an individual object with spatial-temporal information regardless of the difficult situation (low luminance and cluttered backgrounds). The introduced approach is efficient enough to operate tracking in real-time.
Type
conference paper