Detection and tracking of multiple targets in complex environment with an uncalibrated CCD camera is developed in this paper. 1) A background initialization algorithm based on clustering is presented. All stable non-overlapping intervals in the temporal training sequence of each pixel are located as possible backgrounds by slip window; then the background interval is obtained from the classified data set of possible backgrounds by unsupervised clustering. 2) Moving multi-targets are tracked through integration of the motion and shape features by Kalman filter model. In order to ensure the continuity and the stabilization, occlusion processing is performed. The proposed approach is validated under real traffic scenes. Experimental results show that detection and tracking algorithms are robust and adaptive and could be well applied in real-world.
Digital Object Identifier (DOI)
Hu, Hong-Yu; Qu, Zhao-Wei; Li, Zhi-Hui; and Wang, Qing-Nian
"Robust Detection and Tracking Algorithm of Multiple Objects in Complex Scenes,"
Applied Mathematics & Information Sciences: Vol. 08
, Article 45.
Available at: https://dc.naturalspublishing.com/amis/vol08/iss5/45