武大AiFlow无人机视觉测流系统在洪水应急监测中的应用实践
Practical Application of Wuhan University AiFlow UAV Visual Flow Measurement System in Flood Emergency Monitoring

水文监测传统的测流设备往往因地形限制只能进行定点测量，难以满足应急监测需求。本文采用无人机搭载视觉测流仪的方式设计和研制了武大AiFlow无人机视觉测流系统，将先进的视频图像处理技术和改进的STIV算法相结合，实现了在洪水期间流速流量的精准测量。以贵州麦穰水文站、重庆江津五岔水文站以及广西桂平大藤峡为户外实验场地，测试结果表明，与ADCP等常规设备相比，武大AiFlow无人机视觉测流系统在水文监测上的准确度高达90%以上。另外，武大AiFlow无人机视觉测流系统凭借灵活机动、低空巡航的自主飞行方式在数据采集和视频影像传输中展现出了独特优势，不仅能够有效解决洪水期间应急监测的难题，降低高洪期间人工监测的安全风险，还在很大程度上提高了工作效率与测量精度。
Hydrological monitoring of traditional flow measurement equipments are difficult to meet the needs of emergency monitoring, due to terrain and its fixed-point measurement limitations. In this paper, a Wuhan University AirFlow unmanned aerial vehicle (UAV) visual flow measurement system was designed by using a UAV equipped with a visual flow meter, which combined advanced video image processing technology and improved space-time image velocimetry (STIV) algorithm to achieve the accurate measurement of flow velocity and flow rate. Using Mairang, Wucha, and Datengxia stations as outdoor experimental sites, the test results show that the Wuhan University AiFlow UAV visual flow measurement system has an accuracy of more than 90% by compared with the conventional equipment such as acoustic doppler current profiler (ADCP) in hydrological monitoring. In addition, the Wuhan University AiFlow UAV visual flow measurement system shows unique advantages in data collection and video image transmission due to the flexible maneuverability and low-altitude cruising autonomous flight mode, which not only effectively solves the problem of emergency monitoring during floods and reduces the safety risk of manual monitoring during high-flooding periods, but also greatly improves the work efficiency and measurement accuracy.