Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{183981,
        author = {Rukmani Pandey and Prof. Manoj Chaudhary},
        title = {Integrating Traditional Filtering and AI for Single Image Dehazing: A Repeated Averaging Approach},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {3},
        pages = {3899-3909},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=183981},
        abstract = {Image dehazing has emerged as a critical research area in computer vision and digital image processing due to its significant role in enhancing visibility and improving scene understanding under adverse atmospheric conditions. Environmental factors such as fog, mist, dust, and smoke scatter light in the atmosphere, leading to degraded contrast, color distortion, and reduced visual clarity in captured images. This problem directly impacts the performance of vision-based systems in domains such as autonomous driving, unmanned aerial vehicles (UAVs), video surveillance, environmental monitoring, and satellite imaging. Traditional single-image dehazing methods, such as the dark channel prior (DCP) and polarization-based models, are constrained by limitations including halo artifacts, loss of fine details, and excessive computational requirements. Similarly, purely deep learning-based methods often demand large annotated datasets and suffer from overfitting or poor generalization across varied haze densities. To address these challenges, this study proposes a novel AI-powered single-image dehazing framework that combines repeated averaging-based filtering with feedforward neural networks. The repeated averaging filter is employed to estimate ambient illumination and improve radiance consistency, while the neural component refines transmission map estimation and enhances edge preservation. Comprehensive evaluations on standard benchmark datasets reveal that the proposed method consistently outperforms state-of-the-art algorithms in terms of Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), and visual quality. Moreover, the framework demonstrates reduced halo artifacts, improved color restoration, and real-time processing efficiency. These characteristics make the method a robust and scalable solution for next-generation applications in intelligent transportation, drone-based imaging, defense surveillance, and remote sensing systems.},
        keywords = {Image dehazing, single image restoration, artificial intelligence, repeated averaging filter, transmission map, neural network, image enhancement, computer vision, PSNR, SSIM.},
        month = {August},
        }