面向高洪期水流成像测速的目标运动矢量估计方法及试验研究

项目名称： 面向高洪期水流成像测速的目标运动矢量估计方法及试验研究

项目编号： No.61263029

项目类型： 地区科学基金项目

立项/批准年度： 2013

项目学科： 自动化技术、计算机技术

项目作者： 樊棠怀

作者单位： 南昌工程学院

项目金额： 46万元

中文摘要： 高洪期洪水暴涨暴落，尤其是山溪性河流水流速度变化快，含沙量高，漂浮杂物多，致使常规的河流流量测验方案无法开展布置，测速方法和仪器不能正常地施测。针对极端条件下河流流量测验，通过对野外水面复杂光学环境下水流示踪/漂浮物电磁辐射特性分析，研究提出面向水面弱小观测目标特定光谱段及偏振面的光学成像信息获取方法，研究提出仿蜻蜓复眼视觉机制的观测目标检测和运动矢量估计方法，为水面流速测量及流场重建提供可靠基础。探索采用偏振成像和仿生信息处理方法集成创新的手段，以提高观测目标检测和运动矢量估计精度为目标，努力建立一套水流偏振成像测速新的技术方法。通过测量误差控制以及野外现场比测和测验成果精度分析评定，验证和优化偏振成像测速原理、模型和算法，为推进大尺度粒子成像测速（LSPIV）技术方法在野外实用提供理论成果或依据。

中文关键词： 河流流量测验；近红外成像；运动矢量估计；背景抑制；流场定标

英文摘要： In high flood period, water level changes dramatically, especially in mountain rivers. Due to high velocities, high concentration of sediment and floating debris, the regular river flow measuring methods and instruments are highly difficult or impossible to apply to flash floods. For river flow measurement under extreme conditions, two new methods will be proposed and studied based on the analysis of electromagnetic radiation properties of flow tracers / floating debris under complex water surface optical conditions. One is the acquization method of optical imaging Information for weak water surface targets under specific spectral band and the polarization plane, and the other is the target detection and motion vecter estimation method Imitated by the compound eye visual mechanism of dragonfly, to provide a reliable basis for surface velocity measurement and flow field reconstruction. Aimming at improving the accuracy of target detection and motion vector estimation, integrated innovation of the polarization imaging and bionic information processing methods will be explored, and a new flow measurement techniques based on polarization imaging will be established. The proposed principle, model and algorithm of Polarized imaging in flow velocimetry are validated and optimized through measurement error control, file

英文关键词： River flow measurement；NIR imaging；Motion vector estimation；Background suppression；Flow field calibration