View Invariant Human Action Recognition Using Improved Motion Descriptor

Conference paper
Part of the Smart Innovation, Systems and Technologies book series (SIST, volume 33)

Abstract

Human Action Recognition plays a major role in building intelligent surveillance systems. To recognize actions, object tracking and feature extraction phases are crucial to detect the person and track its motion correspondingly. A good motion descriptor is to analyze the motion in each and every frame under different views. As a result, this paper addresses the two major challenges such as motion representation and action recognition. Thus a novel motion descriptor based on optical flow has been proposed to describe the activity of the persons though they are in various views. The new motion descriptor has been framed by fusing shape features and motion features together to increase the performance of the system. This has been fed as the input to the SVM classifier to recognize the human actions. Performance of this system has been tested over Weizmann and IXMAX multi-view data sets and results seem to be promising.

Keywords

Motion descriptor Optical flow Activity recognition 

References

  1. 1.
    Yilmaz, A., Javed, O., Shah, M.: Object tracking :a survey. ACM Comput. Surv. 38(4 Article no. 13), 1–45 (2006)Google Scholar
  2. 2.
    Ryoo, Agarwal, J.K.: Human activity analysis: a survey. ACM Comput. Surv. 43(3), 16:1–16:43 (2011)Google Scholar
  3. 3.
    Poppe, R.: A survey on vision-based human action recognition. Image Vis. Comput. 28, 976–990 (2010)CrossRefGoogle Scholar
  4. 4.
    Gowshikaa, D., Abirami, S., Baskaran, R.: Automated human behaviour analysis from surveillance videos: a survey. Artif. Intell. Rev. (2012). doi: 10.1007/s10462-012-9341-3
  5. 5.
    Gowsikhaa, D., Manjunath, S., Abirami, S.: Suspicious human activity detection from Surveillance videos. Int. J. Internet Distrib. Comput. Syst. 2(2), 141–149 (2012)Google Scholar
  6. 6.
    Sivarathinabala, M., Abirami, S.: Motion tracking of humans under occlusion using blobs. Adv. Comput. Netw. Inf. (Smart Innov. Syst. Technol. 1) 27, 251–258 (2014)Google Scholar
  7. 7.
    Gowsikhaa, D., Abirami, S., Baskaran, R.: Construction of image ontology using low level features for image retrieval. In: Proceedings of the International Conference on Computer Communication and Informatics, pp. 129–134 (2012)Google Scholar
  8. 8.
    Jain, M., J'egou, H., Bouthemy, P.: Better exploiting motion for better action recognition. In: The Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition, pp. 2555–2562 (2013)Google Scholar
  9. 9.
    Huang, K., Wang, S., Tan, T., Maybank, S.: Human behaviour analysis based on new motion descriptor. In: IEEE Transactions on Circuits and Systems for Video Technology (2009)Google Scholar
  10. 10.
    Robertson, N., Reid, I.: A general method for human activity recognition in video. Comput. Vision Image Underst. 104, 232–248 (2006)Google Scholar
  11. 11.
    Sadek, S., Al-Hamadi, A., Michaelis, B., Sayed, U.: A fast statistical approach for human activity recognition. Int. J. Comput. Inf. Syst. Ind. Manag. Appl. 4, 334–340 (2012). ISSN 2150-7988Google Scholar
  12. 12.
    Hu, J., Boulgouris, N.V.: Fast human activity recognition based on structure and motion. Pattern Recogn. Lett. 32, 1814–1821 (2011)CrossRefGoogle Scholar
  13. 13.
    Bruhn, A., Weickert, J.: Lucas/Kanade meets Horn/Schunck: combining local and global optic flow methods. Int. J. Comput. Vision 61(3), 211–231 (2005)CrossRefGoogle Scholar
  14. 14.
  15. 15.

Copyright information

© Springer India 2015

Authors and Affiliations

  1. 1.Department of Information Science and TechnologyAnna UniversityChennaiIndia
  2. 2.Department of Computer Science and Engineering, College of EngineeringAnna UniversityChennaiIndia

Personalised recommendations