Title :
Motion level control in reconstruction of 3D human translation
Author :
Zhang, Li ; Li, Ling
Author_Institution :
Dept. of Comput., Curtin Univ. of Technol., Perth, WA
Abstract :
A motion trend analysis (MTA) system is proposed for 3D human motion reconstruction from uncalibrated monocular video sequences. In this paper, a method named Motion Level Control is proposed to improve the accuracy and efficiency of the MTA system in human translation reconstruction. This method provides automatic assistance in detecting and further amending 3D motion information on transformation of every human body segment, so that the obtained relocation of the human figure represented by the 3D positioning of the pelvis joint can be guaranteed to be reasonable and continuous in 3D space. Various experiments are conducted for synthetic and real video data to demonstrate the functionality and performance of the Motion Level Control method and the motion trend analysis as a whole.
Keywords :
image motion analysis; image reconstruction; image sequences; motion control; video signal processing; 3D human motion reconstruction; 3D human translation reconstruction; 3D positioning; human body segment; motion level control; motion trend analysis; pelvis joint; uncalibrated monocular video sequences; Animation; Cameras; Humans; Image reconstruction; Interference; Level control; Libraries; Motion analysis; Motion pictures; Robotics and automation; Human translation; monocular reconstruction; motion level control;
Conference_Titel :
Control, Automation, Robotics and Vision, 2008. ICARCV 2008. 10th International Conference on
Conference_Location :
Hanoi
Print_ISBN :
978-1-4244-2286-9
Electronic_ISBN :
978-1-4244-2287-6
DOI :
10.1109/ICARCV.2008.4795708
