Abstract. Detecting regions of interest in video sequences is one of the most important tasks in many high level video processing applications. In this paper a novel approach based on support vector data description is presented, which detects foreground regions in videos with quasi-stationary backgrounds. The main contribution of this paper is the novelty detection approach which automatically segments video frames into background/foreground regions. By using support vector data description for each pixel, the decision boundary for the background class is modeled without the need to statistically model its probability density function. The proposed method is able to achieve very accurate foreground region detection rates even in very low contrast video sequences, and in the presence of quasi-stationary backgrounds. As opposed to many statistical background modeling approaches, the only critical parameter that needs to be adjusted in our method is the number of background training frames. 1

Mixture of Gaussians (MOG) has been widely used for robustly modeling complicated backgrounds, especially those with small repetitive movements (such as leaves, bushes, rotating fan, ocean waves, rain). The performance of MOG can be greatly improved by tackling several practical issues. In this paper, we quantitatively evaluate (using the Wallflower benchmarks) the performance of the MOG with and without our modifications. The experimental results show that the MOG, with our modifications, can achieve much better results- even outperforming other state-of-the-art methods. 1.

this paper is organized as follows. In the next section, we discuss previous work on intrinsic images and background generation. In section 3, we explain the computation of the multiplicative background and foreground images. In section 4, we present simulation results and discuss the performance of the proposed method under different temporal scales and illumination conditions

This paper presents a new approach to segmenting monocular videos captured by static or hand-held cameras filming large moving non-rigid foreground objects. The foreground and background objects are modeled using spatialcolor Gaussian mixture models (SCGMM), and segmented using the graph cut algorithm, which minimizes a Markov random field energy function containing the SCGMM models. In view of the existence of a modeling gap between the available SCGMMs and segmentation task of a new frame, one major contribution of our paper is the introduction of a novel foreground/background SCGMM joint tracking algorithm to bridge this space, which greatly improves the segmentation performance in case of complex or rapid motion. Specifically, we propose to combine the two SCGMMs into a generative model of the whole image, and maximize the joint data likelihood using a constrained Expectation-Maximization (EM) algorithm. The effectiveness of the proposed algorithm is demonstrated on a variety of sequences. 1

Abstract — We propose a fast and robust approach to the detection and tracking of moving objects. Our method is based on using lines computed by a gradient-based optical flow and an edge detector. While it is known among researchers that gradientbased optical flow and edges are well matched for accurate computation of velocity, not much attention is paid to creating systems for detecting and tracking objects using this feature. In our method, extracted edges by using optical flow and the edge detector are restored as lines, and background lines of the previous frame are subtracted. Contours of objects are obtained by using snakes to clustered lines. Detected objects are tracked, and each tracked object has a state for handling occlusion and interference. The experimental results on outdoor-scenes show fast and robust performance of our method. The computation time of our method is 0.089 s/frame on a 900 MHz processor. I.

We present a simple, principled approach to detecting foreground objects in video sequences in real-time. Our method is based on an on-line discriminative learning technique that is able to cope with illumination changes due to discontinuous switching, or illumination drifts caused by slower processes such as varying time of the day. Starting from a discriminative learning principle, we derive a training algorithm that, for each pixel, computes a weighted linear combination of selected past observations with timedecay. We present experimental results that show the proposed approach outperforms existing methods on both synthetic sequences and real video data. 1

Detecting regions of interest in video sequences is the most important task in many high level video processing applications. In this paper a robust technique based on recursive learning of video background and foreground models is presented. The proposed modeling technique achieves a fast convergence speed and an adaptive, accurate background/foreground model. Our contributions can be described along four directions. First, a recursive learning scheme is developed to build the models based on colors of the pixels. Our second contribution is to generate background and foreground models to enforce the temporal consistency of detected foregrounds. Third, we exploit dependencies between pixel colors to insure that the model is not restricted to using only independent features. Finally, an adaptive pixel-wise criterion is proposed that incorporates different spatial situations in the scene. We also enforce spatial consistency of the pixels to rule out the effect of erroneously labeled foreground regions. 1.

Abstract not found

Abstract. Background subtraction is an essential task in several static camera based computer vision systems. Background modeling is often challenged by spatio-temporal changes occurring due to local motion and/or variations in illumination conditions. The background model is learned from an image sequence in a number of stages, viz. preprocessing, pixel/region feature extraction and statistical modeling of feature distribution. A number of algorithms, mainly focusing on feature extraction and statistical modeling have been proposed to handle the problems and comparatively little exploration has occurred at the preprocessing stage. Motivated by the fact that disturbances caused by local motions disappear at lower resolutions, we propose to represent the images at multiple scales in the preprocessing stage to learn a pyramid of background models at different resolutions. During operation, foreground pixels are detected first only at the lowest resolution, and only these pixels are further analyzed at higher resolutions to obtain a precise silhouette of the entire foreground blob. Such a scheme is also found to yield a significant reduction in computation. The second contribution in this paper involves the use of the co-linearity statistic (introduced by Mester et al. for the purpose of illumination independent change detection in consecutive frames) as a pixel neighborhood feature by assuming a linear model with a signal modulation factor and additive noise. The use of co-linearity statistic as a feature has shown significant performance improvement over intensity or combined intensity-gradient features. Experimental results and performance comparisons (ROC curves) for the proposed approach with other algorithms show significant improvements for several test sequences. 1

Abstract—Most statistical background subtraction techniques are based on the analysis of temporal color/intensity distribution. However, learning statistics on a series of time frames can be problematic, especially when no frame absent of moving objects is available or when the available memory is not sufficient to store the series of frames needed for learning. In this letter, we propose a spatial variation to the traditional temporal framework. The proposed framework allows statistical motion detection with methods trained on one background frame instead of a series of frames as is usually the case. Our framework includes two spatial background subtraction approaches suitable for different applications. The first approach is meant for scenes having a nonstatic background due to noise, camera jitter or animation in the scene (e.g.,waving trees, fluttering leaves). This approach models each pixel with two PDFs: one unimodal PDF and one multimodal PDF, both trained on one background frame. In this way, the method can handle backgrounds with static and nonstatic areas. The second spatial approach is designed to use as little processing time and memory as possible. Based on the assumption that neighboring pixels often share similar temporal distribution, this second approach models the background with one global mixture of Gaussians. Index Terms—Background detection, motion detection. I.