Substation Isolation Switch State Recognition Technology Based on Image Line Segment Fitting

Abstract

More and more robotics and image processing technologies are applied in smart power substations. This paper deals with the isolation switch state recognition technology based on image processing. Firstly, the substation patrol robot with intelligent image identification system is introduced, and the image acquirement and main processing flowchart including image line segment fitting of substation isolation switch state is presented. Secondly, main algorithms, such as mage graying, image filtering, edge detection and Hough transform, used in the image procedure are provided. Finally, the proposed substation isolation switch state recognition technology is implemented with real image of typical isolating switch, and the experimental results show that the proposed technology can identify the state of the substation isolation switch with accuracy as high as 97%.