面向水下自主作业的环境安全性态势评估

Environmental Security Situation Assessment for Underwater Autonomous Operations

针对水下机器人自主作业安全性态势评估的问题,本文提出一种基于图像的水下环境安全性评估方法。通过单目摄像机采集图像,提取图像中环境物体与作业主体距离、水下物体运动状况和海底地质复杂度等环境要素信息,实现信息融合获得态势评估特征,构建态势评估模型,并对评估结果进行等级量化,得到海洋环境安全状态。此外,为了提高环境态势评估的计算效率,本文运用图像金字塔和四叉树对态势评估算法进行优化。实验表明,态势评估算法能够快速、准确地判断水下环境是否适宜开展自主作业,为精细化水下自主作业提供服务。

The complex marine environment makes underwater autonomous operations face many challenges.Underwater robots need to conduct situation assessments of the marine environment to improve the accuracy of autonomous decision-making.Aiming at the problem of autonomous operation safety situation assessment of underwater robots,this paper proposes an image-based underwater environment safety assessment method.The monocular camera is used to collect images,extract the environmental element information such as the distance between the environmental objects and the main body of the operation,the motion status of the underwater objects,and the complexity of the seabed geology,and realize the information fusion to obtain the situation assessment features,build the situation assessment model,and analyze the assessment results.Level quantification to obtain marine environmental safety status indicators.In addition,to improve the calculation efficiency of environmental situation assessment,this paper uses image pyramid and quadtree to optimize the situation assessment algorithm.Experiments show that the situation assessment algorithm can quickly and accurately judge whether the underwater environment is suitable for autonomous operations,provide services for refined underwater autonomous operations,and play an important role in supporting ocean development.