Maximum correntropy criterion based 3D head tracking with commodity depth camera

3D head tracking becomes easier with the depth image from Microsoft Kinect. However, the noise from face occlusion and illumination still affects the tracking quality. In this paper, we introduce the robust Maximum Correntropy Criterion (MCC) to the problem of 3D head tracking, to tackle these noise...

Full description

Saved in:
Bibliographic Details
Main Authors: Shaoguo Liu, Haibo Wang, Ying Wang, Jixia Zhang, Chunhong Pan
Format: Conference Proceeding
Language:English
Subjects:
3D Head Tracking
Kinect
MCC
Two-Stage Optimization
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:3D head tracking becomes easier with the depth image from Microsoft Kinect. However, the noise from face occlusion and illumination still affects the tracking quality. In this paper, we introduce the robust Maximum Correntropy Criterion (MCC) to the problem of 3D head tracking, to tackle these noises. Fortunately, MCC can handle arbitrarily distributed noises. To solve the MCC based cost function, we develop an effective two-stage optimization scheme with the half-quadric technology. A head tracking system that uses Miscrosoft Kinect is also developed based on the MCC formulation. The system is fully automatic and online, without need of offline training. Experimental results show that the system is very robust against partial occlusion, large motion and sudden illumination variations.
ISSN:1522-4880
2381-8549
DOI:10.1109/ICIP.2013.6738573