基于二进制编码水尺和数字图像的水位远程监控系统设计

发布时间：2018-03-08 23:02

本文选题：二进制编码水尺　切入点：图像处理　出处：《江西理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：水位信息是水库管理过程中最重要的水文数据之一,随着科技与经济的发展,水资源安全利用和优化管理越来越受到社会的关注。目前,传统的水位检测有:(1)人工目测水尺读数。这种方法主观性很强,并且耗费精力,如果遇到极端天气,观测者的人身安全得不到保证。(2)利用水位计采集水位信息。由于水位计在获取水位的过程中,需要专门配置水位信息获取设备,投资大且维护困难。利用数字图像处理技术进行水位识别已经成为研究热点,这种非接触式水位检测系统具有稳定、可靠和实时性强等优点。因此,基于水库监控摄像头的非接触式水位监控系统是一个很有研究价值的课题,这对水库的自动化管理,具有十分重要的意义。本文主要研究内容为:首先,设计了基于二进制编码水尺和数字图像的水位远程监控系统的结构。系统由两个子系统构成,分别是水库水位监测子系统和水库闸门控制子系统,并详细介绍了子系统的组成成分。针对水库水位监测子系统,本文设计了一种二进制编码水尺,并建立了二进制编码水尺的水位识别算法,识别算法采用模块化的思想,主要包含水尺刻度线提取模块、水尺量程信息获取模块、水位值实时解算模块。其次,设计了二进制编码水尺的刻度线提取与字符分割的方案。首先通过视频监控摄像头拍摄含水位信息的二进制编码水尺影像数据,并将影像数据传输给PC机,然后在PC机利用数字图像处理技术进行水尺图像关键像素截取、Radon变换算法的水尺倾斜校正、水尺左右和上下边缘确定等步骤实现水尺刻度线的提取,最后根据坐标的对应关系,实现表征水尺量程的第一个二进制编码字符定位与分割。然后,建立了基于Hopfield神经网络的二进制编码水尺水位识别方法。首先利用离散型Hopfield神经网络的联想记忆能力,对残缺、变形和含噪的二进制编码字符进行修正。然后对修正后的二进制编码字符,利用模板匹配算法实现字符识别,根据二进制编码字符与十进制数字的对应关系,得到水尺的量程信息。最后根据水尺刻度线与水尺量程之间的数学关系,实现水位值的实时解算。最后,分析了水库闸门控制子系统的结构,设计了水位远程监控系统的人机交互界面。针对水库闸门控制子系统,本文利用PLC-变频器技术进行水库闸门的升降控制。本文利用Lab VIEW编程平台实现了水位远程监控系统人机交互界面的设计,通过视频监控摄像头采集实时的水尺图像,并提取水尺图像感兴趣区域,结合本文的水位识别思想,实现水位值的实时解算。通过相应的仿真实验,证明了本文的研究思路是可取的。
[Abstract]:Water level information is one of the most important hydrological data in the process of reservoir management. With the development of science and technology and economy, the safe utilization and optimal management of water resources have attracted more and more attention of the society. The traditional water level measurement includes 1: 1) manual visual measurement of water gauge readings. This method is highly subjective and energy consuming, and if extreme weather is encountered, The safety of the observer is not guaranteed. (2) the water level information is collected by using the water level meter. Since the water level meter needs to be specially equipped with the water level information acquisition equipment during the process of obtaining the water level, Using digital image processing technology to identify water level has become a research hotspot. This non-contact water level detection system has the advantages of stability, reliability and real-time. The contactless water level monitoring system based on reservoir monitoring camera is a valuable research subject, which is of great significance to the automatic management of reservoir. The main contents of this paper are as follows: first of all, The structure of water level remote monitoring system based on binary coded water ruler and digital image is designed. The system is composed of two subsystems: reservoir water level monitoring subsystem and reservoir gate control subsystem. The composition of the subsystem is introduced in detail. For the monitoring subsystem of reservoir water level, this paper designs a binary coded water gauge, and establishes the water level recognition algorithm of binary coded water gauge. The recognition algorithm adopts the idea of modularization. It mainly includes the water scale drawing module, the water gauge range information acquisition module, the real-time water level calculation module. Secondly, In this paper, a scheme of drawing the scale and dividing the characters of the binary coded water ruler is designed. Firstly, the video surveillance camera is used to capture the image data of the binary coded water ruler, which contains the water level information, and the image data is transmitted to the PC. Then the digital image processing technology is used in PC to intercept the key pixels of the water scale image to correct the tilt of the water ruler, to determine the left and right sides of the water ruler and to determine the upper and lower edges of the water ruler. Finally, according to the corresponding relationship of the coordinates, the drawing line of the water ruler is obtained. In this paper, the first binary coded character location and segmentation is realized. Then, a binary coded water level recognition method based on Hopfield neural network is established. Firstly, the associative memory ability of discrete Hopfield neural network is used to deformity. Then the modified binary coded characters are modified by template matching algorithm. According to the corresponding relation between binary coded characters and decimal numbers, Finally, according to the mathematical relationship between the water gauge scale and the water gauge range, the real-time calculation of the water level value is realized. Finally, the structure of the reservoir gate control subsystem is analyzed. The man-machine interface of the remote water level monitoring system is designed. In this paper, the PLC inverter technology is used to control the reservoir gate, and the man-machine interface of the remote water level monitoring system is designed by using Lab VIEW programming platform. The real-time water ruler images are collected by the video surveillance camera. The region of interest of the water ruler image is extracted and the real-time calculation of the water level value is realized by combining the idea of water level recognition in this paper. The corresponding simulation experiment proves that the research idea of this paper is desirable.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP277;TP391.41;TV697

【参考文献】