A method for extracting a moving object in a video by an object detection system operated by at least one processor is provided. The method comprises receiving a certain unsmoothed frame included in the video, and if the certain frame is not a first frame of the video, calculating a pixel intensity difference between reference pixels of a reference background sample extracted from the first frame of the video and each of frame pixels of the certain frame. The method further comprises determining a segmentation of a background and a foreground in the certain frame based on the calculated pixel intensity difference and a predetermined threshold, and extracting the foreground from the certain frame as the object according to a similarity measurement value between the reference pixels and the frame pixels.