Abstract:Subtracting the backgrounds from the video frames is an important step for many video analysis applications. Assuming that the backgrounds are low-rank and the foregrounds are sparse, the robust principle component analysis (RPCA)-based methods have shown promising results. However, the RPCA-based methods suffered from the scale issue, i.e., the ℓ1-sparsity regularizer fails to model the varying sparsity of the moving objects. While several efforts have been made to address this issue with advanced sparse models, previous methods cannot fully exploit the spatial-temporal correlations among the foregrounds. In this paper, we proposed a novel spatial-temporal Gaussian scale mixture (STGSM) model for foreground estimation. In the proposed STGSM model, a temporal consistent constraint is imposed over the estimated foregrounds through nonzero-means Gaussian models. Specifically, the estimates of the foregrounds obtained in the previous frame are used as the prior for these of the current frame, and nonzero means Gaussian scale mixture models (GSM) are developed. To better characterize the temporal correlations, the optical flow has been used to model the correspondences between foreground pixels in adjacent frames. The spatial correlations have also been exploited by considering that local correlated pixels should be characterized by the same STGSM model, leading to further performance improvements. Experimental results on real video datasets show that the proposed method performs comparably or even better than current state-of-the-art background subtraction methods.

Spatiotemporal background modeling based on adaptive mixture of Gaussians

Background modeling using mixture of Gaussians and Laplacian pyramid decomposition.

Adaptive Spatiotemporal Background Modelling

Multi-Channel Adaptive Mixture Background Model for Real-time Tracking.

An Improved Gaussian Mixture Model for an Adaptive Background Model

An Improved Gaussian Mixture Model for Background Estimation

An improved adaptive background modeling algorithm based on Gaussian Mixture Model

Adaptive Background Model for Arbitrary-Long Stationary Target.

Adaptive Learning Algorithm Based on Mixture Gaussian Background

A Novel Adaptive Gaussian Mixture Model for Background Subtraction.

Mixture of Gussian Background Model Based on Walsh-hadamard Transformation

A Double-Subspace Adaptive Background Modeling Method Based on Gaussian Mixture Model

Background subtraction in dynamic scenes with adaptive spatial fusing.

Future-data Driven Modeling of Complex Backgrounds Using Mixture of Gaussians

Spatial-Temporal Gaussian Scale Mixture Modeling for Foreground Estimation

A multiscale spatio-temporal background model for motion detection

Adaptive Background Update Based on Gaussian Mixture Model under Complex Condition

Real-time Tracking in Image Sequences based-on Parameters Updating with Temporal and Spatial Neighborhoods Mixture Gaussian Model

Robust Dynamic Background Model with Adaptive Region Based on T2FS and GMM.

Spatiotemporal GMM for Background Subtraction with Superpixel Hierarchy.

Video background subtracion using improved Adaptive-K Gaussian Mixture Model