基于机载视觉的内河落水人员发现概率建模

Discovery probability modeling of inland waterfall personnel based on airborne vision

为提高无人机对内河落水人员的搜寻发现概率,搭建内河水上搜救海事无人机系统.分析无人机飞行高度与相机视场和相机视觉成像的关系,并结合视觉注意理论,研究搜寻人员在地面站图传显示屏上搜寻落水人员的规律,建立无人机最大航速模型和落水人员发现概率模型.长江实测实验表明,无人机最大航速模型与落水人员发现概率模型设计可靠,可提高内河落水人员搜寻效率.

In order to improve the searching and discovering probability of inland river falling personnel by unmanned aerial vehicle(UAV),a maritime UAV system for inland water searching and rescuing was built.The relationship between UAV flight altitude and camera view field and camera visual imaging was analyzed,and by combining with the theory of visual attention,the regularity of searching people falling into water on the map display screen of ground stations for searchers was studied,and the maximum speed model of UAV and the discovery probability model of dropping personnel were established.The experimental results on the Yangtze River show that the maximum speed model of UAV and the probability model of dropping personnel are reliable for improving the efficiency of inland waterfall personnel.