![]() A larger input image always leads to larger feature maps and greater computational complexity. However, most I2I-based methods suffer from the contradiction between image quality and processing speed, especially when the input image is large. As one of the most popular areas of deep learning, GAN-based image-to-image (I2I) methods provide a general framework for underwater image enhancement ( Islam et al., 2020). Due to the success of deep learning, data-driven approaches provide a potential solution for underwater image enhancement ( Li et al., 2020 Naik et al., 2021). However, these approaches always suffer from heavy color distortion and noise caused by scattering and absorption ( Lu et al., 2017). Some early underwater image enhancement approaches, such as CLAHE ( Pizer et al., 1990), can provide a fast and real-time solution. 256×256) images in real time, so designing a practical underwater enhancement model for real-time large-size underwater image processing is still challenging. Although there has been some progress in the study of underwater image enhancement ( Fabbri et al., 2018 Islam et al., 2020 Naik et al., 2021), these models can only process small-size ( e.g. However, the absorption and scattering of underwater light cause problems such as low contrast, color deviation, and blurred details, seriously affecting the performance of further vision tasks such as exploration, intelligent analysis or subsea operations. ROVs use large-size ( e.g., 960p (1280 × 960) ( Goodman, 2003) underwater vision data to perform the above engineering tasks ( Jenkyns et al., 2015). In recent years, underwater robots such as remotely operated vehicles (ROVs) have been widely used in important tasks such as deep-sea exploration ( Whitcomb et al., 2000), marine species migration, coral reef monitoring ( Shkurti et al., 2012) and the salvage of sunken ships. Enhancement experiments on many real underwater datasets demonstrate our model's advanced performance and improved efficiency. We design our model with equal upsampling blocks (EUBs), equal downsampling blocks (EDBs) and lightweight residual channel attention blocks (RCABs), effectively simplifying the network structure and solving the spatial inconsistency problem. We propose a novel efficiency model, FSpiral-GAN, based on a generative adversarial framework for large-size underwater image enhancement to solve these problems. Furthermore, GAN-based methods tend to generate spatially inconsistent styles that decrease the enhanced image quality. However, most deep learning-based enhancement methods are computationally expensive, restricting their application in real-time large-size underwater image processing. Along with the development of deep learning, underwater image enhancement has made remarkable progress. ![]() ![]() Underwater image enhancement is a fundamental requirement in the field of underwater vision. 4Cofoe Medical Technology Company Limited, Shenzhen, China. ![]() 3School of Life Science and Technology, Xidian University, Xi’an, China.2Key Laboratory of Ocean Observation and Information of Hainan Province, Sanya Oceanographic Institution, Ocean University of China, Sanya, China.1Faculty of Information Science and Engineering, Ocean University of China, Qingdao, China. ![]() Yang Guan 1, Xiaoyan Liu 1, Zhibin Yu 1,2*, Yubo Wang 3, Xingyu Zheng 4, Shaoda Zhang 4* and Bing Zheng 1,2 ![]()
0 Comments
Leave a Reply. |