An improved Camshift based on intensity adaptive compensation

The distribution of color histogram was directly affected by environment illumination. Therefore tracking effect of Camshift algorithm based on color histogram was directly affected. In this paper a self-adaptive compensation method was presented. This method could realize the self-adaptive compensation of awaiting detection image. In view of computational complexity of the camshaft algorithm target search, pyramid down and up sampling and PCA of color histogram were employed to locate the region of target. This method could avoid searching the target in the whole image. This method was testified that it not only could track target but also reduce the computational burden. At last the method which combined illumination self-adaptive compensation and improved Camshift could enhance the robustness of target tracking.