A Moving Objects Detection Method with Resistance to Illumination Change

Abstract

Moving objects detection is conducted in the sequential image of moving objects, which is favorable to detect, identify and analyze the moving objects. It has been applied in video surveillance, virtual reality, and advanced user interfaces. Based on existing research on the Frame Difference Method (FDM) and the Background Subtraction Method (BSM), considering the short time interval between adjacent images used for difference, FDM is adopted for its smaller impact by scene illumination variation, which is complementary to the drawback that BSM is sensitive to environmental variation; while BSM can detect the integral moving objects, which can also make up the disadvantage of FDM in failing to detect the integral moving objects. In this paper, we propose a moving objects detection method with resistance to illumination change. We conclude from the experiment that this method is noise-proof and can adapt the abrupt change in illumination to ensure accuracy of the detection. DOI : http://dx.doi.org/10.11591/telkomnika.v12i5.5094