DocumentCode :
3543429
Title :
Image recognition for on-line vibration monitoring system of transmission line
Author :
Pan, Lizhi ; Xiao, Xiaohui
Author_Institution :
Power & Mech. Eng., Wuhan Univ., Wuhan, China
fYear :
2009
fDate :
16-19 Aug. 2009
Abstract :
Image-based on-line monitoring system is widely applied in China recently, of which the image recognition is one of the key techniques. Considering the image feature and contour characteristic of the transmission line and such accessories as the damper, insulator chains, a recognition method based on chain code tracking and corner point detection is proposed. First extract the edge contour of the components after the image preprocessing, including graying, linear grey transforming upon the original image. Then trace chain code and detect corner points of the edge contour. Finally extract circle contour and determine its center point for the damper end face. Recognition experiment performed on fifty images proves that the proposed method is effective for detecting almost all corner points on image contour and recognizing the objects with 99% reliability.
Keywords :
condition monitoring; edge detection; feature extraction; image recognition; power engineering computing; power overhead lines; vibrations; China; chain code tracking; contour characteristic; corner point detection; edge contour extraction; image feature; image preprocessing; image recognition; linear grey transformation; on-line vibration monitoring system; power transmission lines; Cables; Cameras; Condition monitoring; Image edge detection; Image recognition; Instruments; Insulation life; Pixel; Power transmission lines; Transmission lines; Image Recognition; on-line monitoring system; transmission line;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Electronic Measurement & Instruments, 2009. ICEMI '09. 9th International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4244-3863-1
Electronic_ISBN :
978-1-4244-3864-8
Type :
conf
DOI :
10.1109/ICEMI.2009.5274379
Filename :
5274379
Link To Document :
بازگشت