基于视频数据的水位测量关键技术研究及应用

发布时间：2018-05-06 10:42

本文选题：视频图像 + Hough变换　；参考：《西安理工大学》2016年硕士论文

【摘要】：我国是一个常年经受洪水威胁的国家,随着当前科技的迅速发展,我们在防汛方面已经有了很多突破性的研究工作,然而在一些险峻的地势,建立完整功能的测站是较为困难的,想要较为全面的获取到各个地区的水位数据,图像分析是一个重要的辅助手段。本文通过图像分析手段,对远程获取的水位图像做了一系列的具体研究,其中主要包括了两种远程获取水位数据方法的优化与研究;防汛预警平台的设计与实现。论文主要研究工作如下:首先,本文主要针对水尺图像的水位获取方法进行改进,其中主要包括了水尺图像的预处理部分的水尺细化算法的优化和通过Hough变换的方式进行水尺图像的刻度提取算法的优化,且通过具体的实验证实了改进部分的可实施性。也说明了改进之后的算法在时间复杂度和空间复杂度方面有一定的优势。其次,本文通过视频图像中河道的地理特殊性,提出了一种从河道的水域面积和水位高度之间的关系求取水位数据的新方法。本方法首先通过边缘检测的方式对河岸线进行提取,通过特征值匹配的方式求取具体的河岸线,进而计算出河岸线以下的区域面积,对区域面积和水位高度进行关系拟合来计算动态水位的具体高度。最后,本文对防汛预警平台进行了系统性设计与功能性验证。该系统模块运行状态良好,系统稳定,能够快速且较为准确的为防汛工作人员提供较好的方案依据,得出的结果具有一定的辅助决策价值。
[Abstract]:China is a country that has been threatened by floods for many years. With the rapid development of current science and technology, we have already made a lot of breakthrough research work in flood control. However, in some steep terrain, It is difficult to set up the station with complete function. In order to obtain the water level data of every region comprehensively, image analysis is an important assistant method. Through image analysis, this paper makes a series of concrete research on remotely acquired water level image, including the optimization and research of two remote water level data acquisition methods, and the design and implementation of flood prevention and warning platform. The main research work of this paper is as follows: first, this paper mainly aims at improving the water level acquisition method of water scale image. It mainly includes the optimization of the water scale thinning algorithm in the preprocessing part of the water scale image and the optimization of the scale extraction algorithm of the water ruler image by Hough transform. The implementation of the improved part is proved by the concrete experiments. It also shows that the improved algorithm has some advantages in time complexity and space complexity. Secondly, this paper proposes a new method to obtain the water level data from the relationship between the water area and the water level height through the geographical particularity of the channel in the video image. In this method, the river bank is extracted by edge detection, and the specific river bank is obtained by matching the eigenvalues, and then the area below the river bank is calculated. The specific height of dynamic water level is calculated by fitting the relation between area area and water level height. Finally, this paper carries on the systematic design and the function verification to the flood prevention early warning platform. The system module is running well and the system is stable. The system can provide a better scheme basis for flood control staff quickly and accurately. The result has certain value of auxiliary decision.
【学位授予单位】：西安理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TV123;TP391.41

【参考文献】