水下机器人海底管道巡检数据集

港口海底管道巡检数据集（Port Pipeline Inspection Dataset）聚焦港口及码头附近海底管道的日常巡检需求，主要面向输油管线、供水管线、排污管线等近岸管道设施。数据由搭载高清摄像机、深度相机、声呐设备的水下机器人在港区附近水域实地采集获得。
采集区域以码头作业区、靠泊区、防波堤附近及港池内部为主，经过时空同步、畸变校正、水下去雾增强、降噪处理及人工标注后，形成约 50小时多模态同步数据。

The Port Subsea Pipeline Inspection Dataset focuses on the daily inspection needs of subsea pipelines near ports and wharves, and is primarily developed for nearshore pipeline facilities such as oil transmission pipelines, water supply pipelines, and sewage discharge pipelines. The data was collected on-site in the waters around the port area by underwater robots equipped with high-definition cameras, depth cameras, and sonar equipment. The main collection areas cover wharf operation zones, berthing areas, regions adjacent to breakwaters, and the interior of port basins. After undergoing spatiotemporal synchronization, distortion correction, underwater dehazing enhancement, noise reduction processing, and manual annotation, approximately 50 hours of multimodal synchronized data are finally formed.

搜集汇总

背景与挑战

背景概述

该数据集聚焦港口及码头附近海底管道的日常巡检需求，主要面向输油管线、供水管线、排污管线等近岸管道设施。数据由搭载高清摄像机、深度相机、声呐设备的水下机器人在港区水域实地采集，经处理后形成约50小时多模态同步数据，适用于水下目标检测、管道识别及智能巡检系统开发。