Experiential Sampling based Foreground/Background Segmentation for Video Surveillance

Abstract

Segmentation of foreground and background has been an important research problem arising out of many applications including video surveillance. A method commonly used for segmentation is "background subtraction" or thresholding the difference between the estimated background image and current image. Adaptive Gaussian mixture based background modelling has been proposed by many researchers for increasing the robustness against environmental changes. However, all these methods, being computationally intensive, need to be optimized for efficient and real-time performance especially at a higher image resolution. In this paper, we propose an improved foreground/background segmentation method which uses Experiential Sampling technique to restrict the computational efforts in the region of interest. We exploit the fact that the region of interest in general is present only in a small part of the image, therefore, the attention should only be focused in those regions. The proposed method shows a significant gain in processing speed at the expense of minor loss in accuracy. We provide experimental results and detailed analysis to show the utility of our method.

Background Modeling using Mixture of Gaussians for Foreground Detection - A Survey

“…Indeed, this region of interest in general is present only in a small part of the image, therefore, the attention should only be in these regions. Results presented by Atrey et al [125] show a significant gain in processing speed with a minor loss in accuracy.…”

“…To reduce computation time, Atrey et al [125] propose to use an Experiential Sampling technique. The goal is to apply only the MOG on the region for interest and thus restrict the computational efforts in these regions.…”

“…Computation Time (CT) [24,43,92,[124][125][126][127][128][129][130][131] Memory Requirement (MR) [127,128] time. To solve this problem, Zivkovic [94] proposes an online algorithm that estimates the parameters of the MOG and simultaneously selects the number of Gaussians using the Dirichlet prior.…”

Adaptive background modeling in multicamera system for real-time object detection

“…In Ref. 25, the region of interest has been identified by an experiential sampling technique system that has to be continuously reinitialized in order to detect new objects in the scene. In Ref.…”

EGMM: An enhanced Gaussian mixture model for detecting moving objects with intermittent stops

“…Owing to the outstanding performance of GMM on moving object detection, a number of extended methods based on GMM have been proposed [2,3,4,5]. However, existing GMM based methods can not detect the still objects from moving state for a while (in this paper, we can also call this instance as detecting moving objects with intermittent stops), so it will affect many detection based applications.…”