vioft2nntf2t|tblJournal|Abstract_paper|0xf4ffd0bb03000000f10e000001000500 Monitoring the behavior and activities of people in Video surveillance has gained more applications in Computer vision. This paper proposes a new approach to model the human body in 2D view for the activity analysis using Thinning algorithm. The first step of this work is Background subtraction which is achieved by the frame differencing algorithm. Thinning algorithm has been used to find the skeleton of the human body. After thinning, the thirteen feature points like terminating points, intersecting points, shoulder, elbow, and knee points have been extracted. Here, this research work attempts to represent the body model in three different ways such as Stick figure model, Patch model and Rectangle body model. The activities of humans have been analyzed with the help of 2D model for the pre-defined poses from the monocular video data. Finally, the time consumption and efficiency of our proposed algorithm have been evaluated.
K. Srinivasan1, K. Porkumaran2, G. Sainarayanan3
Sri Ramakrishna Engineering College, India1, Dr. N.G.P. Institute of Technology, India2, ICT Academy of Tamil Nadu, India3
Human Pose Analysis, 2D Human Body Modeling, Pre-processing, Frame Differencing Algorithm, Thinning Algorithm
| January | February | March | April | May | June | July | August | September | October | November | December |
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Published By :
ICTACT
Published In :
ICTACT Journal on Image and Video Processing
( Volume: 1 , Issue: 2 , Pages: 80 - 86 )
Date of Publication :
November 2010
Page Views :
188
Full Text Views :
1
|