Research On The Improved LCD Positioning And Segmentation Algorithm In The Power Meter Reading System

With the popularization of smart grid and the rapid progress of urban modern construction,it is the general trend for people to use efficient meter reading system to ensure the quality of energy consumption monitoring.There are many shortcomings in traditional manual meter reading,such as poor real-time performance and low accuracy,etc,and the limitations of common remote meter reading methods in practical application,such as communication module failure or communication interface can not be opened to the outside world,etc.All of these seriously affect the progress of smart grid construction in China.In order to solve the above problems,this paper designs a new remote meter reading system which is not restricted by the types of meters and communication interfaces.By studying the characteristics of digital display area and digital image recognition technology of intelligent meters,a meter reading system based on image recognition technology is designed,which realizes the functions of image acquisition,LCD positioning,character recognition,remote meter reading center design,information management,etc.Because there are many interference factors in the environment of the meter,the key and difficult point of the system design lies in the research of LCD positioning and character recognition algorithm,which directly determines the overall performance of the meter reading system.The main work of this paper is as follows:(1)In the algorithm of image processing,the whole process is divided into two parts: meter reading area location and character recognition.The location algorithm uses the location screening mode.Firstly,the color edge joint method is used for coarse location of LCD area,and then the results of coarse location are sent to the classification model for precise location.In addition,in the character recognition part,the improved Otsu is firstly used for binarization,character segmentation is carried out according to the characteristics of the character information of the LCD screen,feature extraction is carried out for different characters such as Chinese characters,numbers and graphics,and finally template matching is carried out based on the character characteristics to achieve the recognition purpose.(2)In the meter reading system,the third generation B + of raspberry pie is used as the terminal embedded controllern order to reduce the light interference,the camera is equipped with a light supplement device,and the Wi Fi template on the raspberrypie board is used for communication transmission.The remote center management system is designed in B/S mode.The whole intelligent meter reading system has low cost,reliable performance,practical,easy to use,humanized and other characteristics,and is suitable for a variety of installation environments.The experimental results show that the method can effectively solve the problem of automatic location and recognition of instrument reading area under complex background,improve the recognition accuracy of characters,and the system has high timeliness and strong generalization ability.