Colorogram: A Color Feature Descriptor for Human Blob Labeling

  • Vejey Subash Gandyer
Conference paper
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 222)


Tracking is one of the most important research fields for computer vision. Human Identity tracking is significant in the field of surveillance where a foolproof system is required that not only finds an existence of some person behaving suspicious but also his identification. In this paper, a novel feature descriptor by the name Colorogram has been proposed for labeling human blobs in videos. Two constraints were introduced into the system by way of color information and spatial spread of pixels in a new coordinate system. This feature descriptor helped in identifying/labeling human blobs with a priori knowledge of the scene. Two color spaces (HSV, RGB) and two histogram distance measures (Intersection, Euclidean) were considered for computation. The proposed Colorograms were compared with the conventional Histograms and Spatiograms. Experiments were conducted and the results have been tabulated. It was inferred that the new feature descriptor had classified humans with an acceptable precision and recall rates.


Human identity Colorogram Tracking Feature extraction Histogram Blob Spatiogram 


  1. 1.
    Elgammal A, Harwood D, Davis L (2000) Non-parametric model for background subtraction. In: Proceedings of the 6th European conference on computer vision (ECCV 2000). Dublin, Ireland, Springer, pp 751–767Google Scholar
  2. 2.
    Lipton AJ, Fujiyoshi H, Patil RS (1998) Moving target classification and tracking from real-time video. In: Proceedings of the DARPA image understanding workshop (IUW’98), November 1998. Monterey, USA, pp 129–136Google Scholar
  3. 3.
    Desa SM, Salih QA (2004) Image subtraction in real time moving object extraction. In: Proceedings of international conference on computer graphics, imaging and visualization, January 2004Google Scholar
  4. 4.
    Capellades MB, DeMenthon D, Doermann D (2004) An appearance based approach for consistent labeling of humans and objects in video. In: Proceedings of the pattern analysis and applications, November 2004. pp 1433–7541Google Scholar
  5. 5.
    Chleq N, Br′emond F, Thonnat M (1998) Image understanding for prevention of vandalism in metro stations. In: Proceedings of advanced video-based surveillance systems, November 1998, vol 488. USA, pp 106–116Google Scholar
  6. 6.
    Haritaoglu I, Harwood D, Davis LS (2000) W4: real-time surveillance of people and their actions. IEEE Trans Pattern Anal Mach Intell 22(8):809–830CrossRefGoogle Scholar
  7. 7.
    Alamdar F, Keyvanpour M (2011) A new color feature extraction method based on quad histogram. In: Proceedings of the 3rd international conference on environmental science and information application technology, January 2011, vol 10. pp 777–783Google Scholar
  8. 8.
    Yin X, Duntsch I, Gediga G (2011) Quad tree representation and compression of spatial data. Lect Notes Comput Sci 6499:207–239CrossRefGoogle Scholar
  9. 9.
    Birchfield ST, Rangarajan S (2005) Spatiograms versus histograms for region-based tracking. In: Proceedings of the IEEE conference on computer vision pattern recognition, vol 2. pp 1158–1163Google Scholar
  10. 10.
    Boesen K et al. (2004) Quantitative comparisons of four brain extraction algorithms. Neuro Image 22(3):1255–1261Google Scholar
  11. 11.
    Br′emond F, Thonnat M (1997) Tracking multiple non-rigid objects in a cluttered scene. In: Proceedings of the 10th Scandinavian conference on image analysis (SCIA’97), June 1997, vol. 2. Lappeenranta, Finland, pp 643–650Google Scholar
  12. 12.
    Martinez-Uso A, Salgues G, Velastin SA (2011) Evaluation of unsupervised segmentation algorithms for silhouette extraction in human action video sequences. Lect Notes Comput Sci 7066:13–22CrossRefGoogle Scholar

Copyright information

© Springer India 2013

Authors and Affiliations

  1. 1.Department of Computer Science and EngineeringKCG College of Technology, Anna UniversityChennaiIndia

Personalised recommendations