Robust localisation of pedestrians with cast shadows using homology in a monocular view

In this paper an object detection algorithm is proposed, which is robust in the presence of cast shadows and is based on geometric projections. The novelty of the work lies in the use of homology mapping of the foreground regions between different parallel planes within a monocular view, unlike some existing algorithms which depend on the use of multiple cameras. The results on an open video dataset are provided.

[1]  Kazunori Onoguchi,et al.  Shadow elimination method for moving object detection , 1998, Proceedings. Fourteenth International Conference on Pattern Recognition (Cat. No.98EX170).

[2]  Nicolas Martel-Brisson,et al.  Learning and Removing Cast Shadows through a Multidistribution Approach , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  Ian D. Reid,et al.  Single View Metrology , 2000, International Journal of Computer Vision.

[4]  Jeremy S. Smith,et al.  Real-time detection via homography mapping of foreground polygons from multiple cameras , 2011, 2011 18th IEEE International Conference on Image Processing.

[5]  Mubarak Shah,et al.  Tracking Multiple Occluding People by Localizing on Multiple Scene Planes , 2009, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[6]  Pascal Fua,et al.  Robust Multi-View Change Detection , 2007, BMVC.

[7]  Yael Moses,et al.  Tracking in a Dense Crowd Using Multiple Cameras , 2010, International Journal of Computer Vision.

[8]  W. Eric L. Grimson,et al.  Adaptive background mixture models for real-time tracking , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).