Title: A new approach for human activity analysis through identification of body parts using skin colour segmentation

Authors: K. Srinivasan, K. Porkumaran, G. Sainarayanan

Addresses: Department of Electronics and Instrumentation Engineering, Sri Ramakrishna Engineering College, Coimbatore 641 022, India. ' Dr. N.G.P. Institute of Technology, Coimbatore 641 022, India. ' Department of Research and Development, ICT Academy of Tamil Nadu, Chennai 600 096, India

Abstract: The objective of this paper is to give a new approach for identifying the human body parts like head, hands and legs for activity analysis. Here, the proposal is made for using different colour space algorithms for identifying the elements of the human body. We have used six dissimilar colour space algorithms such as RGB, YCbCr, HSV1, HSV2, HSI and rgb (Normalised RGB) to analyse the human activities in indoor video sequences. Here, 11 activities have been considered. In the proposed work, the RGB skin model provides the efficiency of 96.35% whereas YCbCr colour space algorithm gives 45.51%.

Keywords: skin colour segmentation; colour space models; video surveillance; activity analysis; human activity; body parts; body part identification; indoor video sequences; RGB skin modelling; video processing; human motion analysis.

DOI: 10.1504/IJSISE.2010.034998

International Journal of Signal and Imaging Systems Engineering, 2010 Vol.3 No.2, pp.93 - 104

Received: 11 Sep 2009
Accepted: 26 May 2010
Published online: 31 Aug 2010 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article