虚拟水下数据集

虚拟水下数据集是由纽卡斯尔大学工程学院创建的，旨在为自主检查提供水下环境的数据。该数据集包含550张图像，这些图像是通过将实验室环境中的物体图像与水下环境图像结合生成的。数据集的创建过程利用了生成对抗网络（GANs），将实验室图像转换为具有水下特征的图像，以模拟真实水下环境。这些人工数据集可用于水下物体图像分类和检测，以增强水下操作，解决因难以获取真实水下图像而产生的问题。

The Virtual Underwater Dataset was developed by the School of Engineering, Newcastle University, with the goal of providing underwater environment data for autonomous inspection. This dataset includes 550 images generated by combining images of objects captured in laboratory settings with underwater scene images. The dataset creation process leverages Generative Adversarial Networks (GANs) to convert laboratory images into ones with authentic underwater features, thereby simulating real underwater environments. This artificial dataset can be applied to underwater object image classification and detection, to enhance underwater operations, and address the challenges arising from the difficulty of acquiring real underwater images.