图像法测流系统开发与应用

Development and application of an image-based flow measurement system

基于市面上现有的视频监控摄像机开发一套图像法测流系统,实现非接触式的水尺水位、表面流速和断面流量远程在线监测。测流系统采用风光互补供电,并运用虚拟局域网(VPN)技术和4G移动通信网络实现远程访问,可在野外自治运行长达7个阴雨天。通过集成少量低成本的辅助测量设备及对测量算法在现场应用中的优化设计,能够有效抑制水面耀光、阴影等噪声的干扰,并可直接利用天然漂浮物或水面模式进行表面流速测量。测试结果表明,测流系统能够在数分钟内完成一次测量,得到现场的视频图像、稳定的时均水位、测点高度密集的流速分布,以及子断面高达上百个的断面流量。相比传统测流仪器具有时空分辨率高的优势,有利于捕获暴雨、山洪等极端条件下的水文过程。

Using the existing video surveillance camera, an image-based flow measurement system is developed to realize on-line telemetry of non-contact water level, surface velocity and discharge. The system is powered by a wind and photovoltaic hybrid power source and remotely accessed via the 4 G mobile communication network with VPN technology. It can automatically run in the wild for up to 7 rainy days. A small number of low-cost auxiliary measuring devices are integrated and the measuring algorithms are optimized for field application. On account of this, the influence of water surface noises like glares and shadows is effectively decreased, and the surface velocities can be directly measured by tracing of nature floating debris or free-surface patterns. Testing results show that, the system can perform single measurement in several minutes, output data including video and image on site, stable timeaveraged water level, high density velocity distribution and discharge with subsections up to hundreds. Compared with traditional flow measurement instruments, it is superior in high spatial and temporal resolution, thus has great potential to record the hydrological processes under extreme conditions, such as rainstorm and flash flood.