Image Editing Based Data Augmentation for Illumination-Insensitive Background Subtraction

Dimitrios Sakkos, Edmond S. L. Ho, Hubert P. H. Shum and Garry Elvin
Journal of Enterprise Information Management (JEIM), 2020

REF 2021 Submitted Output Impact Factor: 6.5 Top 25% Journal in Information Science & Library Science

Image Editing Based Data Augmentation for Illumination-Insensitive Background Subtraction

Abstract

A core challenge in background subtraction (BGS) is handling videos with sudden illumination changes in consecutive frames. While the use of data augmentation has been shown to increase robustness, the modelling of realistic illumination changes remains less explored and is usually limited to global, static brightness adjustments. In this paper, we focus on tackling the problem of background subtraction using augmented training data, and propose an augmentation method which vastly improves the model’s performance under challenging illumination conditions. In particular, our framework consists of a local component that considers direct light/shadow and lighting angles, and a global component that considers the overall contrast, sharpness and color saturation of the image. It generates realistic, structured training data with different illumination conditions, enabling our deep learning system to be trained effectively for background subtraction even when significant illumination changes take place. We further propose a post-processing method that removes noise from the output binary map of segmentation, resulting in a cleaner, more accurate segmentation map that can generalise to multiple scenes of different conditions. Experimental results demonstrate that the proposed system outperforms existing work, with the highest F-measure score of 81.27% obtained by the full system. To facilitate the research in the field, we open the source code of this project at: https://github.com/dksakkos/illumination_augmentation

Downloads

YouTube

Citations

BibTeX

@article{sakkos20image,
 author={Sakkos, Dimitrios and Ho, Edmond S. L. and Shum, Hubert P. H. and Elvin, Garry},
 journal={Journal of Enterprise Information Management},
 title={Image Editing Based Data Augmentation for Illumination-Insensitive Background Subtraction},
 year={2020},
 doi={10.1108/JEIM-02-2020-0042},
 issn={1741-0398},
 publisher={Emerald Publishing Limited},
}

RIS

TY  - JOUR
AU  - Sakkos, Dimitrios
AU  - Ho, Edmond S. L.
AU  - Shum, Hubert P. H.
AU  - Elvin, Garry
T2  - Journal of Enterprise Information Management
TI  - Image Editing Based Data Augmentation for Illumination-Insensitive Background Subtraction
PY  - 2020
DO  - 10.1108/JEIM-02-2020-0042
SN  - 1741-0398
PB  - Emerald Publishing Limited
ER  - 

Plain Text

Dimitrios Sakkos, Edmond S. L. Ho, Hubert P. H. Shum and Garry Elvin, "Image Editing Based Data Augmentation for Illumination-Insensitive Background Subtraction," Journal of Enterprise Information Management, Emerald Publishing Limited, 2020.

Supporting Grants

Similar Research

Dimitrios Sakkos, Hubert P. H. Shum and Edmond S. L. Ho, "Illumination-Based Data Augmentation for Robust Background Subtraction", Proceedings of the 2019 International Conference on Software, Knowledge, Information Management and Applications (SKIMA), 2019
Dimitrios Sakkos, Edmond S. L. Ho and Hubert P. H. Shum, "Illumination-Aware Multi-Task GANs for Foreground Segmentation", IEEE Access, 2019

 

 

Last updated on 14 March 2024
RSS Feed