Review and Evaluation of Commonly-Implemented Background Subtraction Algorithms

Abstract

Locating moving objects in a video sequence is the first step of many computer vision applications. Among the various motion-detection techniques, background subtraction methods are commonly implemented, especially for applications relying on a fixed camera. Since the basic inter-frame difference with global threshold is often a too simplistic method, more elaborate (and often probabilistic) methods have been proposed. These methods often aim at making the detection process more robust to noise, background motion and camera jitter. In this paper, we present commonly-implemented background subtraction algorithms and we evaluate them quantitatively. In order to gauge performances of each method, tests are performed on a wide range of real, synthetic and semi-synthetic video sequences representing different challenges. * This work was realized with the financial help of the Regional Council of Le Centre and the French Industry Ministry within the Capthom project of the Competitiveness Pole S 2 E 2

Object Detection and Recognition for Assistive Robots: Experimentation and Implementation

“…With the purpose of validating the obtained qualitative results, a quantitative evaluation has been carried out. In this case, the true positive rate (TPR) and false positive rate (FPR) measurements are used [80]. That is, the proportion of correctly classified positives (TPR); and the proportion of incorrectly classified negatives (FPR).…”

Background Subtraction on Embedded Hardware

“…In practice though, most techniques exhibit a small performance increase for the classification task when using RGB instead of grayscale features [12]. Thus, from a classification perspective and despite that the computation time is more or less tripled, it is beneficial to use color images, even when colors have been interpolated in the image.…”

Improving Detection Results of Differential Background Subtraction Method by Merging it with C-HDC Algorithm

“…On the other hand, if image background slightly changes or input frames are noisy, the best algorithms to use are GMM, KDE and 1-G. A proper choice of distance measure is also very important. The comparison of all algorithms mentioned above and the results obtained can be found in [1]. The research conducted and described in this paper was based on a simple differential BS method.…”