An Illumination Invariant Change Detection Algorithm

In this paper, a homomorphic filtering based change detection algorithm is proposed to detect moving objects from light-varing monocular image sequences. In our approach, a background model is first constructed, and background subtraction is applied to classify image pixels into background or foreground. We utilize illumination invariant local components to model the background, which are obtained using homomorphic filtering. Threshold for every pixel in the image can be selected automatically to accommodate the change of lighting. In addition, the connectivity information is integrated into the background-foreground classification process by Bayesian estimation. Experimental results show that the presented approach works well in the presence of heavy moving shadows and illumination variance.

[1]  H. Niemann,et al.  Adaptive change detection for real-time surveillance applications , 2000, Proceedings Third IEEE International Workshop on Visual Surveillance.

[2]  Touradj Ebrahimi,et al.  Change detection based on color edges , 2001, ISCAS 2001. The 2001 IEEE International Symposium on Circuits and Systems (Cat. No.01CH37196).

[3]  Til Aach,et al.  Change Detection in Image Sequences Using Gibbs Random Fields , 1993 .

[4]  Aaron F. Bobick,et al.  Fast Lighting Independent Background Subtraction , 2004, International Journal of Computer Vision.

[5]  Alex Pentland,et al.  Pfinder: real-time tracking of the human body , 1996, Proceedings of the Second International Conference on Automatic Face and Gesture Recognition.

[6]  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).

[7]  Trevor Darrell,et al.  Background estimation and removal based on range and color , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[8]  Olaf Munkelt,et al.  Adaptive Background Estimation and Foreground Detection using Kalman-Filtering , 1995 .

[9]  Robert C. Bolles,et al.  Background modeling for segmentation of video-rate stereo sequences , 1998, Proceedings. 1998 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No.98CB36231).