Background subtraction methods in video streams: A review

Abstract

Background subtraction is one of the most important parts in image and video processing field. There are some unnecessary parts during the image or video processing, and should be removed, because they lead to more execution time or required memory. Several subtraction methods have been presented for the time being, but find the best-suited method is an issue, which this study is going to address. Furthermore, each process needs to the specific subtraction technique, and knowing this issue helps researchers to achieve faster and higher performance in their research. This paper presents a comparative study of several existing background subtraction methods which have been investigated from simple background subtraction to more complex statistical techniques. The goal of this study is to provide a view of the strengths and drawbacks of the widely used methods. The methods are compared based on their memory requirement, the computational time and their robustness of different videos. Finally, a comparison between the existing methods has been employed with some factors like computational time or memory requirements. It is also hoped that this analysis helps researchers to address the difficulty of selecting the most convenient method for background subtraction.

Comparative study of motion detection methods for video surveillance systems

“…Bouwmans 10 Mixture of Gaussians (MoG) 15 Mixture of Gaussians with particle swarm optimization (MoG-PSO) 100 Improved MoG 101 MoG with MRF 102 MoG improved HLS color space 103 Spatial-Time adaptive per pixel mixture of Gaussian (S-TAP-MoG) 104 Adaptive spatio-temporal neighborhood analysis (ASTNA) 105 Subspace learning-principle component analysis (Eigen-Backgrounds) 42 Subspace learning independent component analysis (SL-ICA) 106 Subspace learning incremental non-negative matrix factorization (SL-INMF) 107 Subspace learning using incremental ranktensor (SL-IRT) 108 Wallflower Bayesian multi-layer 36 Histogram over time 13 Local-self similarity 110 Bianco et al 130 IUTIS-1 130 IUTIS-2 130 IUTIS-3 130 Flux tensor with split Gaussian models (FTSG) 124 Self-balanced local sensitivity (SuBSENSE) 123 Weightless neural networks (CwisarDH) 51,122 Spectral-360 121 fast self-tuning BS 120 K-nearest neighbor method (KNN) 119 Kernel density estimation(KDE) 16 Spatially coherent self-organized background subtraction (SC-SOBS) 22 Fish4knowledge § §143 for underwater fish detection and tracking. Other comparative studies can be found in [144][145][146][147][148] .…”

“…Xu et al 132 Mixture of Gaussians (MoG) 15 Kernel density estimation(KDE) 16 Codebook 19 Self-organized background subtraction (SOBS) 21 ViBe 39 Pixel-based adaptive segmenter (PBAS) 40 GMM Zivkovic 60 (adaptive GMM) sample consensus (SACON) 133,134 CDnet 2014 dataset 125 Video dataset proposed in Wen et al Fish4knowledge § §143 for underwater fish detection and tracking. Other comparative studies can be found in [144][145][146][147][148] .…”

A comprehensive review of vehicle detection using computer vision

“…For surveillance systems, one of the most popular types of devices are visible spectrum cameras, which have been extensively used for monitoring environments, events, activities, and people. Different studies in background-foreground segmentation [4][5][6][7] and object detection classification [8][9][10] have also attempted to automatically analyze image or video data from surveillance cameras.…”

Particle Swarm Optimization and Support Vector Machine for Vehicle Type Classification in Video Stream

“…Background subtraction is a method often used to detect moving objects on a video. Compared to the detection of vehicle objects using a single image, the detection using sequential frames from a video is easier because the foreground movement has significant movement and the background is assumed to be static, so moving vehicle objects on the video frame can be detected from foreground segmentation [5]. Frame Differencing is a method of background subtraction that can be used to detect moving objects effectively, by subtraction the current video frame with the previous video frame, then thresholding process is carried out to get moving objects [4].…”