Font Size: a A A

Identification And Research Of Instrument Pointer Number In Complex Background

Posted on:2018-01-08Degree:MasterType:Thesis
Country:ChinaCandidate:C W YangFull Text:PDF
GTID:2348330515953265Subject:Computer technology
Abstract/Summary:PDF Full Text Request
With the rapid development of electric power intelligent system in our country,the demand for power automation and unattended operation is becoming more and more urgent.The pointer instrument,as a measuring tool to indicate the state of power environment,is widely used in power plants.Because the pointer meter is completed by manual interpretation,it will appear inefficient,error and so on.In view of this phenomenon,this paper studies and designs a pointer indicator recognition system under complex background.The machine is used instead of man to interpret the instrument of power plant,so as to improve the work efficiency and reduce the error rate.In this paper,through the deep understanding of the characteristics and structure of the pointer instrument,combined with the advanced technology in the field of artificial intelligence,the components of the instrument,the instrument pointer and the dial number are analyzed.This paper presents a method of precise positioning instrument based on Adaboost classifier,the method can effectively solve the problem caused by the complicated environment difficult to locate the substation instrument,stability and robustness of the instrument positioning compared with Hough transform has higher.This paper uses Sift algorithm to realize the rotation and tilt dial dial in the number of correct positioning,to achieve the needle extraction by Hough line detection,and according to the calculation formula of plane vector angle to complete the interpretation of a pointer to the current number of shows.In the research and design of pointer meters,according to the detection results of Sift algorithm,this paper proposes a feature region localization method based on clustering algorithm.According to the principle of plane clustering,the method is located in the densest region of Sift matching point and achieves efficient digital positioning of the disk.In the course of the study on the extraction of pointer of instrument,this paper proposes a relocation method based on local pointer center position,the method of using Hough to realize the detection results of re positioning the pointer position by position between the pointer and the center of the dial,and then complete the instrument pointer more accurate extraction.In the design structure,this topic uses hand hold terminal to carry on the scene information collection,the server side carries on the C/S frame system which carries on the collection image information processing.The instrument pointer recognition system has been realized according to the actual demand,and the accurate identification of pointer number of COMPOUND GAUGE(vacuum pressure)instrument has been completed.
Keywords/Search Tags:instrument pointer, Adaboost classifier, Sift feature, Hough line detection, vector angle
PDF Full Text Request
Related items