An Advanced Method for Human Body Detection

Abstract:

An advanced method is proposed that combines the advantage of Hough transform and exemplar method for human detection. This method firstly uses a universal codebook for all the patches in an image which reduces the total codebook size and the dimensionality of similarity features. Another contribution is in classifying and detection, in which section we use the sprite of Hough transform, but we train the model in a boost way. Finally the experiment results are given.

• Periodical: Applied Mechanics and Materials (Volumes 380-384)
• Pages: 3886-3889
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.380-384.3886
• Online since: August 2013
• Authors: Feng Xia Li, San Yuan Zhao, Zhong Jun Li, Wang Cheng Zhang
• Keywords: Dimensionality Reduction, Exemplar Method

References

[1] B. Leibe, A. Leonardis, and B. Schiele. Combined object categorization and segmentation with an implicit shape model. In ECCV'04 Workshop on Stat. Learn. in Comp. Vis., pages 17–32, (2004).

DOI: 10.1007/11957959_26

[2] Y. Shan, F. Han, H. Sawhney, and R. Kumar. Learning Examplar-Based Categorization for the Detection of Multi-View Multi-Pose Objects. CVPR 2006. 2, 7.

DOI: 10.1109/cvpr.2006.168

[3] William Robson Schwartz, Aniruddha Kembhavi, David Harwood, and Larry S. Davis, Human detection using partial least squares analysis, in ICCV, (2009).

DOI: 10.1109/iccv.2009.5459205

[4] Boris Babenko, Piotr Dollar, et al, Simultaneous Learning and Alignment: Multi-Instance and Multi-Pose Learning, ECCV (2008).

[5] Yimeng Zhang, et al, Implicit Shape Kernel for Discriminative Learning of the Hough Transform Detector, BMVC 2010 doi: 10. 5244/C. 24. 105.