Abstract:Background subtraction is one of the most fundamental and challenging tasks in computer vision. Many background subtraction algorithms work well under the assumption that the backgrounds are static over short time periods but degrade dramatically in dynamic scenes, such as swaying trees, rippling water, and waving curtains. In this paper, we propose an effective background subtraction method to address these difficulties by combining color features with texture features in the ViBe framework. Specifically, we present a novel local compact binary count (LCBC) feature that can capture local binary gray-scale difference information and totally discard the local binary structural information. The effective fusion of color and LCBC information significantly improves the performance of the ViBe model, making it very robust to background variations while still highlighting the moving objects. We further embed the total variation (TV) norm regularization technique into the proposed method, which can enhance the spatial smoothness of foreground objects, thereby further improving the accuracy of the method. We evaluate the proposed method against ten sequences containing dynamic backgrounds and show that our method outperforms many state-of-the-art methods in reducing the false positives without compromising the reasonable foreground definitions. The experimental results on challenging well-known data sets demonstrate that the proposed method works effectively on a wide range of dynamic background scenes.

Dynamic background subtraction based on appearance and motion pattern

Dynamic Background Subtraction Based on Spatial Extended Center-Symmetric Local Binary Pattern

Dynamic Background Modeling and Subtraction Using Spatio-Temporal Local Binary Patterns

Dynamic Background Subtraction Using Spatial-Color Binary Patterns

Dynamic Background Subtraction Based on Local Dependency Histogram.

A new background subtraction method using texture and color information

Detecting Moving Objects from Dynamic Background Combining Subspace Learning with Mixed Norm Approach

Hybrid center-symmetric local pattern for dynamic background subtraction

Complex Background Subtraction by Pursuing Dynamic Spatio-Temporal Models

Local Compact Binary Count Based Nonparametric Background Modeling for Foreground Detection in Dynamic Scenes

Texture And Motion Pattern Fusion For Background Subtraction

Spatial-temporal Nonparametric Background Subtraction in Dynamic Scenes

A temporal-spatial background modeling of dynamic scenes

Motion-Based Background Subtraction

Multl-resolution Background Subtraction for Dynamic Scenes

Background Segmentation of Dynamic Scenes Based on Dual Model

Local spatial co-occurrence for background subtraction via adaptive binned kernel estimation

A covariance-based method for dynamic background subtraction

A Fuzzy Background Model For Moving Object Detection

Background subtraction via time continuity and texture consistency constraints.

Dynamic Background Estimation and Complementary Learning for Pixel-Wise Foreground/background Segmentation