Detection of camouflaged moving objects based on Laws energy and varying frequency
Qian Li · Chinese Journal of Stereology and Image Analysis · 2011
The detection of moving objects in dynamic background is a research focus in video surveillance.Detecting camouflaged moving objects is a challenge.The traditional Gaussian mixture model(GMM) is not advantageous for detecting these moving objects.By analyzing the color change of camouflaged moving objects in dynamic background,we put forward an effective method for camouflaged moving objects using energy of Laws and a special statistic quantity called varying frequency.The experiment shows that,when the camera is fixed,the method is effective for detecting camouflaged moving objects in dynamic background when the color of moving objects change greatly or moving objects have complicated texture.