Time Ordering Shuffling for Improving Background Subtraction

Abstract

Abstract. By construction, a video is a series of ordered frames, whose order is dened at the time of the acquisition process. Background subtraction methods then take this input video and produce a series of segmentation maps expressed in terms of foreground objects and scene background. To our knowledge, this natural ordering of frames has never been questioned or challenged. In this paper, we propose to challenge, in a prospective view, the natural ordering of video frames in the context of background subtraction, and examine alternative time orderings. The idea consists in changing the order before background subtraction is applied, by means of shuing strategies, and re-ordering the segmentation maps afterwards. For this purpose, we propose several shuing strategies and show that, for some background subtraction methods, results are preserved or even improved. The practical advantage of time shuing is that it can been applied to any existing background subtraction seamlessly.

LaBGen: A method based on motion detection for generating the background of a scene

“…An odd pass processes the input sequence in the forwards (chronological) order, while an even pass processes it in the backwards (reverse) order to ensure smooth transitions between the end of a pass and the beginning of a new one. In general, considering the input frames in a non-chronological order does not necessarily decrease the performance of background subtraction, and in some cases, it can improve the performance with respect to the forwards order, as shown in Laugraud et al (2015a). In the current implementation, P is always chosen to be odd to finish with a forwards pass.…”