HUMAN ACTIVITY TRACKING FOR WIDE-AREA SURVEILLANCE

We present a method for tracking and identifying moving persons from video images taken by a fixed field-ofview camera. Specifically, we have developed a system to first track moving persons in a given scene and generate color-based models of those persons to accomplish identification. The tracking is non-invasive meaning that it does not require persons to wear any particular electronics or clothing to be able to track them. Tracking is accomplished using a position-based data association algorithm while the color modeling is accomplished using a mixture-ofGaussians statistical model. The expectation-maximization algorithm is used to generate the color models over a sequence of frames of data in order to make the entire process near-realtime. Once a color model is developed for a given person, that model will be placed in a database where it will be used for future identification. Any similar identification scheme can be used in place of color modeling for greater accuracy.

[1]  Narendra Ahuja,et al.  Gaussian mixture model for human skin color and its applications in image and video databases , 1998, Electronic Imaging.

[2]  Jake K. Aggarwal,et al.  Tracking human motion in an indoor environment , 1995, Proceedings., International Conference on Image Processing.

[3]  Gerhard Rigoll,et al.  Person tracking in real-world scenarios using statistical methods , 2000, Proceedings Fourth IEEE International Conference on Automatic Face and Gesture Recognition (Cat. No. PR00580).

[4]  Scott A. Nichols.  IMPROVEMENT OF THE CAMERA CALIBRATION THROUGH THE USE OF MACHINE LEARNING TECHNIQUES , 2001 .

[5]  Roberto Brunelli,et al.  Person identification using multiple cues , 1995, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[6]  Hironobu Fujiyoshi,et al.  Moving target classification and tracking from real-time video , 1998, Proceedings Fourth IEEE Workshop on Applications of Computer Vision. WACV'98 (Cat. No.98EX201).

[7]  E. Brookner.  Tracking and Kalman Filtering Made Easy , 1998 .

[8]  Surendra Ranganath,et al.  Tracking people , 2002, Object recognition supported by user interaction for service robots.

[9]  Mubarak Shah.  Tracking people in presence of occlusion , 2000 .

[10]  J. L. Roux.  An Introduction to the Kalman Filter , 2003 .

[11]  Michael J. Black,et al.  Cardboard people: a parameterized model of articulated image motion , 1996, Proceedings of the Second International Conference on Automatic Face and Gesture Recognition.

[12]  Ioannis Pitas,et al.  Segmentation and tracking of faces in color images , 1996, Proceedings of the Second International Conference on Automatic Face and Gesture Recognition.

[13]  Marc Acheroy,et al.  Automatic Face Authentication from 3D surface , 1998, BMVC.

[14]  EEL6825: Pattern Recognition Maximum-Likelihood Estimation for Mixture Models: the EM algorithm , 2003 .

[15]  Larry S. Davis,et al.  W4S : A real-time system for detecting and tracking people in 2 D , 1998, eccv 1998.

[16]  Alexander H. Waibel,et al.  Skin-Color Modeling and Adaptation , 1998, ACCV.

[17]  Iván R. Zapata.  DETECTING HUMANS IN VIDEO SEQUENCES USING STATISTICAL COLOR AND SHAPE MODELS , 2001 .

[18]  Larry S. Davis,et al.  W/sup 4/: Who? When? Where? What? A real time system for detecting and tracking people , 1998, Proceedings Third IEEE International Conference on Automatic Face and Gesture Recognition.

[19]  Takeo Kanade,et al.  Human Face Detection in Visual Scenes , 1995, NIPS.