Paper Title :A Novel Image Enhancement Framework for ROVS in the Freshwater Environment
Author :Austin Mao, Jiajun Jiang, Chung-Hao Chen, Wen-Chao Yang
Article Citation :Austin Mao ,Jiajun Jiang ,Chung-Hao Chen ,Wen-Chao Yang ,
(2023 ) " A Novel Image Enhancement Framework for ROVS in the Freshwater Environment " ,
International Journal of Mechanical and Production Engineering (IJMPE) ,
pp. 7-13,
Volume-11,Issue-7
Abstract : This paper presents a novel deep learning-based framework for enhancing underwater images. The proposed
framework follows a two-step approach, beginning with haze removal as a preprocessing step, followed by image
enhancement using Water-Net, an advanced deep learning model. According to our comprehensive evaluations using both
publicly and privately collected datasets from lakes and rivers, experimental results show the effectiveness of our method for
improving the quality of underwater images, particularly in challenging scenarios characterized by high turbidity or limited
visibility in freshwater. Our method excels in various tasks including deblurring, contrast enhancement, and color vividness
improvement. This adaptability of our proposed method makes it highly suitable for applications in underwater target detection
and terrain exploration, specifically in the fresh water environment. Furthermore, our proposed method offers an easy
integration framework for existing state-of-the-art techniques, facilitating future advancements in image enhancement needs
for under-water related applications. To our knowledge, very few investigations have been dedicated to the enhancement of the
quality of underwater images in the freshwater environment. The contribution of our work can fill in this knowledge gap.
Keywords - Underwater Image Enhancement, Image Quality Improvement, Reduce Haze, Remotely Operated Vehicles
(ROVs)
Type : Research paper
Published : Volume-11,Issue-7
DOIONLINE NO - IJMPE-IRAJ-DOIONLINE-19990
View Here
Copyright: © Institute of Research and Journals
| |
 |
| |
 |
PDF |
| |
Viewed - 34 |
| |
Published on 2023-11-10 |
|
