Large scale Underwater Image Quality Data set (LUIQD)

LUIQD includes 64,180 real and enhance underwater images covering a wide range of scenes, target and imaging conditions, with their perceptual quality scores. The LUIQD is developed to assess the quality of both real underwater images and those enhanced by various UIE methods. A number of 6,418 raw underwater images showcase a diverse range of scenes and degradation conditions were collected, and a total of 64,180 images are acquired using nine different traditional and advanced deep-learning-based UIE methods.  

LUIQD数据集包含64180张真实水下图像与增强水下图像，涵盖多样场景、拍摄目标与成像环境，并附带对应的感知质量评分。本数据集旨在评估真实水下图像，以及经各类水下图像增强（Underwater Image Enhancement，UIE）方法处理后的图像的质量。我们首先收集了6418张原始水下图像，这些图像涵盖丰富的场景类型与退化条件；随后通过9种不同的传统及基于先进深度学习的水下图像增强方法，对每张原始图像进行增强处理，最终得到总计64180张图像。