Design And Implementation Of Power Equipment Identification System

With the development of image processing and recognition technology,more and more related applications come into being.These applications are based on image processing algorithms to identify and classify things.Because of the many kinds of power equipment,it is very important to test the appearance of the electric power equipment.At present,the inspectors of the electric power academy are using the tools such as the ruler to measure them manually.If there is a recognition system,it can be used to install on a handheld equipment,and you can take photos of the power equipment and identify and detect it by the handheld equipment,which will greatly facilitate the work of the inspectors of the power science academy.In order to solve this problem,this paper designs and implements a power equipment identification and detection system,and transplants the system to the handheld devices of the Android system.This system realizes the function of identify and detect the power equipment by taking photos.This paper first introduces the overall design structure of the power equipment identification system.This system is divided into three modules,which are the UI interface module,the camera module and the recognition algorithm module.In these three modules,each of the required techniques and principles are described in detail.In the part of the designed algorithm module,according to the different characteristics of different equipment,we can use the algorithm to judge the type of equipment,then find the corresponding appearance standard specifications by the type of the equipment,and compare with the equipment,so as to judge whether the equipment is qualified,and then realize the identification and detection of the power equipment.The algorithms used for each feature are as follows.Firstly,Gauss filter is used for image denoising,Hough transformation is used to detect screen edge,then threshold method is used for internal area judgment,and canny contour method is selected for button or keyboard area to detect lines in the area.The method is used to detect lines in the area.Then binarization and histogram statistics were carried out to determine whether there was a button or keyboard in this area.In view of whether there is an indicator,the method of histogram statistics is selected to judge.Finally,we design and implement an algorithm to detect whether these areas are within the range of specifications.The last part of this paper is the test part,we take a lot of photos on network message analysis device,gateway machine,measurement and control device,synchronous vector measurement device and time synchronization device respectively.From these,100 qualified photos of each sort equipment are selected,totally 500 photos are tested.The system has been tested and meets the functional and performance requirements.