ABOShips
收藏资源简介:
ABOShips数据集是由Åbo Akademi大学理学与工程学院创建,包含9880张图像,精确标注了41967个海洋船舶实例。该数据集考虑了多种因素如背景变化、大气条件、光照、可见比例、遮挡和尺度变化,以确保数据集能真实反映海洋环境。数据集内容涵盖9种船舶类型、浮标及其他漂浮物。创建过程中,首先进行初步标注,随后使用CSRT跟踪器检测标注不一致性并重新标注。ABOShips数据集主要应用于海洋交通监控、海洋监视和导航安全等领域,旨在通过机器视觉技术提高船舶检测的准确性和效率。
The ABOShips dataset was created by the Faculty of Science and Engineering at Åbo Akademi University. It consists of 9,880 images with precisely annotated 41,967 marine vessel instances. To ensure the dataset authentically reflects real-world marine environments, multiple influencing factors are taken into account, including background variations, atmospheric conditions, lighting, visibility proportions, occlusion, and scale changes. The dataset covers 9 types of vessels, buoys, and other floating objects. During its creation, preliminary annotations were first conducted, followed by the detection of annotation inconsistencies and subsequent re-annotation using the CSRT tracker. The ABOShips dataset is primarily applied in fields such as marine traffic monitoring, maritime surveillance and navigation safety, aiming to improve the accuracy and efficiency of vessel detection via machine vision technologies.




