Illumination-Based Data Augmentation for Robust Background Subtraction

A core challenge in background subtraction (BGS) is handling videos with sudden illumination changes in consecutive frames. In this paper, we tackle the problem from a data point-of-view using data augmentation. Our method performs data augmentation that not only creates endless data on the fly, but also features semantic transformations of illumination which enhance the generalisation of the model. It successfully simulates flashes and shadows by applying the Euclidean distance transform over a binary mask generated randomly. Such data allows us to effectively train an illumination-invariant deep learning model for BGS. Experimental results demonstrate the contribution of the synthetics in the ability of the models to perform BGS even when significant illumination changes take place.

• Publication year

  2019

• Venue

  International Conference on Software, Knowledge, Information Management and Applications

• Publication date

  2019-06-20

• Fields of study

  Computer Science, Engineering

• Identifiers
  DOI 10.1109/SKIMA47702.2019.8982527arXiv 1910.08470
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Establishing Pose Based Features Using Histograms for the Detection of Abnormal Infant Movements

  2019cited by this paper
• Moving Object Detection in Complex Scene Using Spatiotemporal Structured-Sparse RPCA

  2019cited by this paper
• Fusion-based foreground enhancement for background subtraction using multivariate multi-model Gaussian distribution

  2018cited by this paper
• Background Subtraction Using Spatio-Temporal Group Sparsity Recovery

  2018cited by this paper
• Multiscale Fully Convolutional Network for Foreground Object Detection in Infrared Videos

  2018cited by this paper
• Real-time nonparametric background subtraction with tracking-based foreground update

  2018cited by this paper
• Foreground Segmentation Using a Triplet Convolutional Neural Network for Multiscale Feature Encoding

  2018cited by this paper
• Foreground Segmentation with Tree-Structured Sparse RPCA

  2018cited by this paper
• End-to-end video background subtraction with 3d convolutional neural networks

  2018cited by this paper
• Three-Stream Convolution Networks After Background Subtraction for Action Recognition

  2018cited by this paper
• $L_0$ Regularized Stationary-Time Estimation for Crowd Analysis

  2017cited by this paper
• Improving posture classification accuracy for depth sensor-based human activity monitoring in smart environments

  2016cited by this paper
• Segmentation Models

  2016cited by this paper
• Sensitivity Enhancement in Magnetic Particle Imaging by Background Subtraction

  2016cited by this paper
• Real-Time and Robust Compressive Background Subtraction for Embedded Camera Networks

  2016cited by this paper
• Vehicle tracking in low hue contrast based on CAMShift and background subtraction

  2015cited by this paper
• Modeling spatial relations of human body parts for indexing and retrieving close character interactions

  2015cited by this paper
• PIRM: Fast background subtraction under sudden, local illumination changes via probabilistic illumination range modelling

  2015cited by this paper
• Hand tracking and gesture recognition system for human-computer interaction using low-cost hardware

  2015cited by this paper
• Very Deep Convolutional Networks for Large-Scale Image Recognition

  2014influential reference
• Fully convolutional networks for semantic segmentation

  2014cited by this paper
• Adam: A Method for Stochastic Optimization

  2014cited by this paper
• A comprehensive review of background subtraction algorithms evaluated with synthetic and real videos

  2014cited by this paper
• Symmetry-driven accumulation of local features for human characterization and re-identification

  2013cited by this paper
• ViBe: A Universal Background Subtraction Algorithm for Video Sequences

  2011cited by this paper
• Evaluation of background subtraction techniques for video surveillance

  2011influential reference
• Background Subtraction under Sudden Illumination Changes

  2010cited by this paper
• Making Background Subtraction Robust to Sudden Illumination Changes

  2008cited by this paper
• Efficient adaptive density estimation per image pixel for the task of background subtraction

  2006cited by this paper
• Submitted to Ieee Transactions on Pattern Analysis and Machine Intelligence. Special Section on Video Surveillance and Monitoring a Bayesian Computer Vision System for Modeling Human Interactions

  year unknowncited by this paper

• Morphological Background-Subtraction Modeling for Analyzing Traffic Flow

  2025cites this paper
• Shadow-Robust Semantic Segmentation for Autonomous Navigation in Walking Space

  2024cites this paper
• A Survey of Synthetic Data Augmentation Methods in Machine Vision

  2024cites this paper
• Highlighting Challenges of State-of-the-Art Semantic Segmentation with HAIR - A Dataset of Historical Aerial Images

  2024cites this paper
• Domain Generalization for In-Orbit 6D Pose Estimation

  2024cites this paper
• The Machine Vision in Wire Harness Industry for Fuse Box Inspection

  2024cites this paper
• An experimental evaluation of Siamese Neural Networks for robot localization using omnidirectional imaging in indoor environments

  2024cites this paper
• Domain Generalization for 6D Pose Estimation Through NeRF-based Image Synthesis

  2024cites this paper
• Leveraging Neural Network Uncertainty in Adaptive Unscented Kalman Filter for Spacecraft Pose Estimation

  2023influential citation
• Robust color medical image segmentation on unseen domain by randomized illumination enhancement

  2022cites this paper
• A Feature-space Multimodal Data Augmentation Technique for Text-video Retrieval

  2022cites this paper
• Textile defect detection using YOLOv5 on AITEX Dataset

  2022influential citation
• Random Shadows and Highlights: A new data augmentation method for extreme lighting conditions

  2021cites this paper
• Day to Night Image Style Transfer with Light Control

  2021cites this paper
• Image editing-based data augmentation for illumination-insensitive background subtraction

  2020cites this paper
• Image Editing based Data Augmentation for Illumination-insensitive Background Subtraction

  2020influential citation