FLSea
收藏资源简介:
FLSea数据集是由海法大学的海洋技术系创建,专注于水下视觉-惯性和立体视觉的前视数据。该数据集包含5个立体视觉数据集和8个视觉-惯性数据集,每个数据集包含数千张图像,涵盖不同的水下能见度和环境光条件。数据集的创建过程涉及使用商业摄影测量软件Agisoft Metashape生成离线地面实况深度图。FLSea数据集主要用于开发水下应用,如自主避障、视觉里程计、3D跟踪、同时定位与地图构建(SLAM)和深度估计。
The FLSea dataset was created by the Department of Marine Technology at the University of Haifa, focusing on forward-looking underwater visual-inertial and stereo vision data. It includes 5 stereo vision datasets and 8 visual-inertial datasets, each containing thousands of images covering various underwater visibility and ambient lighting conditions. The dataset creation process involves using the commercial photogrammetry software Agisoft Metashape to generate offline ground-truth depth maps. The FLSea dataset is mainly used for developing underwater applications such as autonomous obstacle avoidance, visual odometry, 3D tracking, simultaneous localization and mapping (SLAM), and depth estimation.

- 1FLSea: Underwater Visual-Inertial and Stereo-Vision Forward-Looking Datasets海洋技术系,海法大学,海法,以色列 · 2023年



