With the view of effectively fitting the complicated water level process of the lower Yellow River, polynomial regression, stepwise regression, parameters by ridge estimate and so on, are logically integrated. And the...With the view of effectively fitting the complicated water level process of the lower Yellow River, polynomial regression, stepwise regression, parameters by ridge estimate and so on, are logically integrated. And the progressive transformation is introduced. Then a new method is put forward. The core difference of this new method from the same kind of methods lies in that in this method the strong coupling effect of weak influencing factors which is common in a complicated water level process is considered, that many effective methods are synthetically used to reduce the fitting model error, and that the necessary progressive transformation is introduced. The advantages of many theories and methods are logically integrated in this method, and the method can be easily used. The rationality and necessity of each step in this method are ensured by sufficient theories, so this method can be widely used to effectively simulate the inherent relations in the same kind of complicated data. Furthermore, many complicated water level processes of the lower Yellow River are fitted by this method, and all the fitting precisions are markedly higher than the precision by the other existing methods. Every component term in the fitting model has clear physical meaning.展开更多
针对库水位传统量测方法中水尺易锈蚀倾斜,精度低且成本高的问题,该文提出一种不依赖水尺,基于图像匹配的库水位变动识别方法。首先对监控相机拍摄的库坝上游面光学图像进行畸变消除和透视变换,消除相机误差。进一步选取包含水面的感兴...针对库水位传统量测方法中水尺易锈蚀倾斜,精度低且成本高的问题,该文提出一种不依赖水尺,基于图像匹配的库水位变动识别方法。首先对监控相机拍摄的库坝上游面光学图像进行畸变消除和透视变换,消除相机误差。进一步选取包含水面的感兴趣区域(region of interest,ROI)图像进行自适应二值化分割、形态学处理等前处理操作,将水面和库坝特征分离,突出水位线位置。最后,利用归一化互相关匹配算法(normalised cross correlation,NCC)对水位变动前后的两幅图像进行匹配计算,识别水位线变动距离。通过室内试验与现场测试验证上述方法实用性。结果表明:基于图像匹配的库水位动态识别方法可准确识别水位变动,相对误差约为5%,此算法鲁棒性较高。该研究可为库水位自动化、低成本监测提供一种新思路。展开更多
基金Supported by Natural Science Fund of Hohai University (Grant No. 2007428611)
文摘With the view of effectively fitting the complicated water level process of the lower Yellow River, polynomial regression, stepwise regression, parameters by ridge estimate and so on, are logically integrated. And the progressive transformation is introduced. Then a new method is put forward. The core difference of this new method from the same kind of methods lies in that in this method the strong coupling effect of weak influencing factors which is common in a complicated water level process is considered, that many effective methods are synthetically used to reduce the fitting model error, and that the necessary progressive transformation is introduced. The advantages of many theories and methods are logically integrated in this method, and the method can be easily used. The rationality and necessity of each step in this method are ensured by sufficient theories, so this method can be widely used to effectively simulate the inherent relations in the same kind of complicated data. Furthermore, many complicated water level processes of the lower Yellow River are fitted by this method, and all the fitting precisions are markedly higher than the precision by the other existing methods. Every component term in the fitting model has clear physical meaning.
文摘针对库水位传统量测方法中水尺易锈蚀倾斜,精度低且成本高的问题,该文提出一种不依赖水尺,基于图像匹配的库水位变动识别方法。首先对监控相机拍摄的库坝上游面光学图像进行畸变消除和透视变换,消除相机误差。进一步选取包含水面的感兴趣区域(region of interest,ROI)图像进行自适应二值化分割、形态学处理等前处理操作,将水面和库坝特征分离,突出水位线位置。最后,利用归一化互相关匹配算法(normalised cross correlation,NCC)对水位变动前后的两幅图像进行匹配计算,识别水位线变动距离。通过室内试验与现场测试验证上述方法实用性。结果表明:基于图像匹配的库水位动态识别方法可准确识别水位变动,相对误差约为5%,此算法鲁棒性较高。该研究可为库水位自动化、低成本监测提供一种新思路。
