|
Anil Kumar Vishwakarma
Reema Ajmera
Dinesh k. Dharamdasani
Keywords:
Image Dehazing, Generative Artificial Intelligence, CMFNet, Remote Sensing, Vegetation Mapping, Underwater Imaging, Deep Learning, PSNR, SSIM, Computer Vision.
Abstract:
Haze significantly degrades image clarity across remote sensing, vegetation mapping, and underwater applications, affecting tasks such as land monitoring, environmental studies, agriculture, and marine research. This paper presents a generative AI-based Channel-wise Multi-scale Feature Fusion Network (CMFNet) for robust image dehazing. The methodology integrates public datasets, systematic pre-processing, supervised learning with optimized training-validation strategies, and evaluation using PSNR, SSIM, Mutual Correlation, and Average Gradient. Results indicate notable improvements in image clarity and detail preservation compared to existing approaches. Ethical practices and limitations are acknowledged, with future work directed toward real-time and multimodal solutions.
|
|
International Journal of Recent Research and Review
ISSN: 2277-8322
Vol. XVII, Issue 3
September 2024
|
PDF View
PUBLISHED
September 2024
ISSUE
Vol. XVII, Issue 3
SECTION
Articles
|
