Real-world Underwater Image Enhancement (RUIE) data set
收藏资源简介:
RUIE数据集是由大连理工大学信息科学与工程学院和辽宁省泛在网络与服务软件重点实验室创建的大型水下图像数据集,包含超过4000张图像,旨在解决水下图像增强中的挑战。数据集分为三个子集,分别针对图像可见性、颜色偏移和高级检测/分类任务。创建过程中,研究人员在黄海的张子岛附近使用22个防水视频摄像头捕捉了超过250小时的水下视频,涵盖了光照、深度、模糊度和颜色偏移的广泛变化。RUIE数据集不仅用于评估和比较水下图像增强算法,还为训练智能水下车辆提供了数据基础,特别是在解决图像质量对高级视觉任务影响的研究中发挥重要作用。
The RUIE dataset is a large-scale underwater image dataset developed by the School of Information Science and Engineering, Dalian University of Technology and the Key Laboratory of Ubiquitous Networks and Service Software of Liaoning Province. Consisting of over 4000 images, it is designed to address the challenges in underwater image enhancement. The dataset is divided into three subsets targeting image visibility, color cast, and advanced detection/classification tasks respectively. During its development, researchers captured over 250 hours of underwater footage using 22 waterproof video cameras near Zhangzi Island in the Yellow Sea, covering a wide range of variations in illumination, depth, blurriness, and color cast. The RUIE dataset not only serves as a benchmark for evaluating and comparing underwater image enhancement algorithms, but also provides a data foundation for training intelligent underwater vehicles, playing a critical role in research investigating the impact of image quality on advanced visual tasks.

- 1Real-world Underwater Enhancement: Challenges, Benchmarks, and Solutions大连理工大学信息科学与工程学院和辽宁省泛在网络与服务软件重点实验室 · 2019年



