Multiple Kernel Tracking with SSD

Kernel-based objective functions optimized using the mean shift algorithm have been demonstrated as an effective means of tracking in video sequences. The resulting algorithms combine the robustness and invariance properties afforded by traditional density-based measures of image similarity, while connecting these techniques to continuous optimization algorithms. This paper demonstrates a connection between kernelbased algorithms and more traditional template tracking methods. There is a well known equivalence between the kernel-based objective function and an SSD-like measure on kernel-modulated histograms. It is shown that under suitable conditions, the SSD-like measure can be optimized using Newton-style iterations. This method of optimization is more efficient (requires fewer steps to converge) than mean shift and makes fewer assumptions on the form of the underlying kernel structure. In addition, the methods naturally extend to objective functions optimizing more elaborate parame...

• Type: Conference
• Year: 2004
• Where: CVPR
• Authors: Gregory D. Hager, Maneesh Dewan, Charles V. Stewart
• Keywords: Computer Vision, Continuous Optimization Algorithms, CVPR 2004, Kernel-based Objective Functions, Mean Shift Algorithm, Objective Functions, Template Tracking Methods