摘要
为解决进行海洋浮体运动测量时测量速度与测量精度无法同时保障的问题,通过将YOLO目标检测算法与双目深度相机相结合,提出了一种基于机器视觉的浮体运动测量系统,通过振动台实验与波浪水槽进行了验证实验.结果表明:该系统可准确稳定地实时追踪与测量快速运动浮体的空间位置,并且适用于真实的波浪运动环境.研究成果为海洋工程实践中的浮体测量问题提供了更加优良的解决方案,进而有助于提升海上人工作业的安全性.
In the report,in order to solve the problem that the measuring speed and the measuring precision can not be guaranteed at the same time,based on the combination of YOLO target detection algorithm and stereo depth cameras,a novel approach for motion measurement of marine buoys based on machine vision was pro⁃posed,and a fixed-type vibration table experiment was performed for validating.The results indicated that the system can accurately and stably track and measure the spatial position of the buoy marker,and which can be used in the real environment.Our findings provide a better solution for the real-time tracking and measurement of marine buoys,and which is helpful for improving the safety of manual work at sea.
作者
张文厂
陈龙
李延巍
任兴月
Zhang Wenchang;Chen Long;Li Yanwei;Ren Xingyue(College of Civil Engineering and Architectural,Hainan University,Haikou 570228,China)
出处
《海南大学学报(自然科学版)》
CAS
2024年第3期340-346,共7页
Journal of Hainan University(Natural Science)
基金
国家重点研发计划项目(RZ2300001800)。
关键词
海洋浮体
追踪与测量
机器视觉
双目深度相机
深度学习
marine floater
tracking and measurement
machine vision
stereo depth camera
deep learning