Font Size: a A A

Research And Application Of Digital Image Processing Technology In Hydrological Monitoring System

Posted on:2018-11-18Degree:MasterType:Thesis
Country:ChinaCandidate:F T ShenFull Text:PDF
GTID:2310330536957296Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
China has a vast territory.Because of numerous waters,the flood disasters happen frequently,which not only caused economic losses,also make people suffer a lot.In order to solve these problems,using hydrological monitoring data information to locate and regulate the water resources reasonably is essential.Design intelligent automatic hydrological monitoring systems by cutting-edge technology has great significance to the country's economic construction and social stability.Measurement of the water level and the flow rate is the main work of hydrological monitoring.Ccurrent measurement methods are based on data transmission by remote sensors,where there are some limitations in the prophase preparation and real time data collection.With the rise of machine vision system in recent years,combined with the embedded system,a convenient,economical and efficient remote monitoring system can be designed,which measure the water level and flow rate with image processing and pattern recognition.In this paper,a new method of water level measurement is studied.The method is aimed at the waters which have complex environment,where the traditional water level meter is not easy to install,but with hydrological rulers.This method uses video transmission technology to transmit the real-time image of rivers to the main control room.Recognizing and segmenting the water ruler automatically,reading the hydrology ruler scale line,determining the intersection of surface and hydrological scale lines,so as to determine the real-time water level.Color image processing method is used in this paper,according to the color characteristic of the hydrological ruler,converting the RGB color space to the YCrCb color space.After extracting the component,using the global adaptive threshold binarization,morphology processing,edge detection and Hough transform to segment the water ruler.Using the histogram statistics,skeleton extraction and other methods to read the scale value of the hydrological ruler and determine the real-time water level finally.The research in this paper can provide a remote unattended monitoring system to the hydrological monitoring.The preparatory work is simple and convenient,reducing the manpower cost,especially suitable for remote and poor water environment.Provide strong support for the monitoring and invoking of water resources,increase the science and technology of remote hydrological monitoring.
Keywords/Search Tags:water level monitoring, hydrological ruler characteristics, color image processing, image segmentation, adaptive threshold
PDF Full Text Request
Related items