Deep Image Deraining
Laila Hegazy, Shereen Moataz Afifi, Omar M. Fahmy · 2023
Image restoration is a crucial area within computer vision, with the goal of recovering high-quality images from degraded ones. To achieve the best outcomes, a delicate balance between intricate spatial details and comprehensive contextual information is essential. This research focuses on image deraining, which is a critical domain in computer vision for recovering high-quality images from degraded observations. The study explores various deraining methods through a comprehensive literature review, highlighting their strengths and weaknesses. The centerpiece is the MPRNet architecture, an innovative approach for image deraining, which is thoroughly studied and examined. The research also proposes a methodology to optimize the MPRNet model's performance and improve the quality of restored images resulting in better results both qualitatively and quantitatively. Overall, the research contributes to the field by introducing a novel approach and optimizing the MPRNet model's performance, paving the way for future advancements in image restoration.