Moving objects detection based on kernel independent component analysis

In video moving objects detection, the same illumination and perspective, will lead to that moving objects and background is nonlinearly mixed. In this paper, Kernel Independent Component Analysis (KICA) algorithm is proposed to detect the video moving objects. By the selection of kernel's parameters, the proposed algorithm avoids the unreasonable assumption which moving objects image and background image is completely independent in the video moving objects detection based on the Independent Component Analysis (ICA). And it achieves the separation of moving objects and background in the feature space by kernel function. Experimental results demonstrate that KICA is better than ICA in video motion detection. (C) 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of [CEIS 2011]