The Research Of Icing Detection For Transmission Lines Based On Image-Processing

As China’s economy continues to accelerate the pace of construction and the powerload increases every year, the reliability of grid becomes increasingly important. Thetransmission line icing largely caused some threat to the power grid reliability and safeoperation. In order to reduce the insufficiency of the current transmission line icingdetection, this thesis presents a highly automated transmission line icing detection methodbased on image processing.This thesis analyzed the causes, influencing factors, types and characteristics oftransmission line icing, and introduced an active contour model based image segmentationmethod. Based on this thesis proposed an overall program of image processing based icingdetection of transmission line, this program can automatically capture, process andanalyze an image. The results will be sent back to the control dispatch center by using awireless communication network.This thesis focuses on the image process method of ice on transmission lines and thecalculation of ice thickness. The thesis firstly introduced the icing image acquisitionparameters, and detailed image’s gray scale, enhancement, filtering, and pre-process space.Subsequently, this thesis describes the use of traditional methods of image segmentationand active contour model based image segmentation methods, including classical Snakemodel, and the steps of level set based image segmentation methods. Meanwhile, thesemethods were tested, compared and analyzed by the results.Then, the thesis presents a calculation method of power line ice thickness, includingthe calibration of imaging module and analysis of transmission lines outline of the imageto calculate ice thickness. And then describes the steps to verify the test method and thismethod was tested to verify the feasibility.