OUC dataset
收藏资源简介:
OUC数据集是由一组研究者构建的大型水下图像数据集,旨在研究水下图像增强算法对水下目标检测任务的影响。该数据集包含4400张水下图像,每张图像都有相应的参考图像和边界框注释。这些图像在不同的光照和浑浊度条件下拍摄,具有多种水下图像特征,如颜色偏移、对比度降低和雾霾水平。数据集的创建过程中,研究者采用了一种结合主观感知和客观评估的新型混合参考图像生成方法,以确保参考图像的质量。OUC数据集主要用于评估和比较不同的水下图像增强算法,以及研究这些算法如何影响水下目标检测的性能,从而为水下机器人和海洋工程等领域的应用提供支持。
The OUC Dataset is a large-scale underwater image dataset developed by a team of researchers, aiming to investigate the impact of underwater image enhancement algorithms on underwater object detection tasks. This dataset contains 4,400 underwater images, each paired with corresponding reference images and bounding box annotations. These images were captured under varying lighting and turbidity conditions, exhibiting typical underwater image artifacts such as color cast, reduced contrast, and haze-like distortions. During the dataset construction process, the researchers adopted a novel hybrid reference image generation method that combines subjective perception and objective evaluation to ensure the quality of the reference images. The OUC Dataset is primarily used to evaluate and compare different underwater image enhancement algorithms, as well as to study how these algorithms affect the performance of underwater object detection, thereby providing support for applications in fields such as underwater robotics and marine engineering.

- 1A Benchmark dataset for both underwater image enhancement and underwater object detection未提及 · 2020年



