Abstract
This chapter proposes a fully automatic framework for real-time human head pose estimation from a monocular image based on geometric approach. Our method starts with Constraint Local Model-based (CLM-based) facial feature tracking. Combined with tracked facial feature locations and a statistical 3D human face model pre-captured from Kinect device, 3D translation and 3D orientation of head pose are estimated by Pose from Orthography and Scaling with Iterations (POSIT) algorithm. Since both CLM-based facial feature tracking and POSIT algorithm are fast, our method can achieve real-time head pose estimation with more than 15 frames/s (fps) in an Intel i5-3230 CPU 3.0 GHz processor without GPU acceleration. By using Kinect’s head pose estimation results as ground truth data, our estimation results show that head location and orientations can achieve tracking accuracy within 2 cm and 5° standard deviations, respectively.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
