A Novel Bilateral Recurrent Network Approach for Robust Rain Streak Removal

DOI

10.2991/978-94-6463-716-8_14How to use a DOI?

Keywords

CNN; Machine Learning; BRN; BLSTM; Deep-Learning

Abstract

Rain streaks in outdoor images are a challenging problem for computer vision applications, including but not limited to autonomous driving, surveillance video monitoring, and remote sensing. Streaks in the image tend to reduce its quality and, as a result, can affect identifying, tracking, and determining objects. Traditional methods for image deraining are not suitable for real-world applications because of sub-optimal performance. Even with the optimization-based methods and crafted priors, the traditional approach often removes the rain streaks & image details. Practically, majority methods fail to differentiate rain straights from unadulterated images, which leads to image degradation and loss of detail. Image-aware deraining methods provide improved accuracy and image quality. Convolutional Neural Networks (CNNs) are effective in single-image rain streak removal using deep learning mechanisms. However, the methods of Deep Neural Network (DNN) often struggle to accurately model real-world conditions, as they tend to obscure the rain patterns against the background. Moreover, the traditional methods have high computational costs. This paper solves the single-image deraining problem by introducing the Bilateral Recurrent Network (BRN). The BRN integrates recurrent with the Bilateral Long-Short Memory cells to use the temporal and pattern information of the rain patterns. Experimental results on the datasets show BRN exceeds prior-dated works in single-image deraining. This report comprehensively evaluates the BRN model, implementation, and experimental evaluation, showing that the model is a solid, real-world solution for single image deraining.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Ch. Rathan Kumar
AU  - T. Sunil Kumar
AU  - Ajeet K. Jain
PY  - 2025
DA  - 2025/05/26
TI  - A Novel Bilateral Recurrent Network Approach for Robust Rain Streak Removal
BT  - Proceedings of the International Conference on Recent Advancements and Modernisations in Sustainable Intelligent Technologies and Applications (RAMSITA 2025)
PB  - Atlantis Press
SP  - 159
EP  - 172
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-716-8_14
DO  - 10.2991/978-94-6463-716-8_14
ID  - Kumar2025
ER  -