综合影像白平衡的稀疏光流跟踪河流测速方法  

Sparse Optical Flow Tracking Method for River Velocity Measurement Based on Image White Balance

作　　者：刘雅楠 李好胜 邓刚 胡娜 LIU Ya′nan;LI Haosheng;DENG Gang;HU Na(School of Geography and Information Engineering,China University of Geosciences(Wuhan),Wuhan 430074,China;Wuhan Zhitu Yunqi Technology Co.,Ltd.,Wuhan 430074,China)

机构地区：[1]中国地质大学(武汉)地理与信息工程学院,湖北武汉430074 [2]武汉智图云起科技有限公司,湖北武汉430074

出　　处：《测绘与空间地理信息》2023年第8期81-84,共4页Geomatics & Spatial Information Technology

基　　金：赣江流域水文气象遥感多源数据协同水资源评价项目(2020110021)资助。

摘　　要：为了提高天然河流测速的实际应用能力和测量精度,本文提出了一种综合影像白平衡的稀疏光流跟踪河流测速方法。实验证明:本方法未经图像预处理时的测量精度与大尺度粒子图像测速(large-scale particle image velocimetry,LSPIV)相当,然而,本方法无需人工辅助示踪,测量成本低,测量时间减少了约50%,显著提高了测量效率,且对水面复杂的成像条件具有更好的鲁棒性,在天然示踪场景下的应用能力更优。在预处理阶段引入基于影像自动白平衡(auto white balance,AWB)的全反射算法。在高速流场下,测量精度有了显著提升。本方法测量精度合理,人工输入参量少,计算效率和自动化程度高,在高速流场下有一定的应用前景。In order to improve the practical application ability and measurement accuracy of natural river velocity measurement,a sparse optical flow tracking river velocity measurement method based on integrated image white balance is proposed in this paper.Experiments show that:(1) The measurement accuracy of this method without image preprocessing is equivalent to that of large scale particle image velocimetry(LSPIV).However,this method does not require manual tracing,has low measurement cost,reduces measurement time by about 50%,significantly improves measurement efficiency,and has better robustness to complex imaging conditions of water surface,and has better application ability in natural tracing scenes.(2) In the pre-processing stage,the automatic white balance(AWB) based on image is introduced,which significantly improves the measurement accuracy in high-speed flow fields.This method has reasonable measurement accuracy,less manual input parameters,high calculation efficiency and automation,and has certain application prospects in high-speed flow field.

关 键 词：河流测速 稀疏光流 自动白平衡 测量效率 

分 类 号：P25[天文地球—测绘科学与技术] TP391[自动化与计算机技术—计算机应用技术]