Method for detecting icing thickness of power transmission line insulator

Abstract

The invention relates to a method for detecting the icing thickness of a power transmission line insulator. The method comprises the following steps: acquiring a current insulator image; acquiring foreground frame information in the insulator image through a template matching algorithm, and performing foreground and background segmentation through a GrabCut algorithm in combination with the foreground frame information; performing dimension reduction on the image data based on the segmentation result so as to reduce the edge sensitivity of the image, and therefore, an insulator contour can besearched and drawn in the original image; estimating the actual icing thickness after the drawn insulator contour area is calculated, and if the estimated actual icing thickness exceeds a preset threshold value, sending out a hazard early warning signal. The method can automatically and accurately obtain the optimal foreground frame, can effectively improve the efficiency of the whole image processing flow, and facilitates the real-time automatic monitoring and early warning of the power transmission line.

Description

technical field [0001] The invention relates to the technical field of transmission line maintenance, in particular to a method for detecting the thickness of ice coating on an insulator of a transmission line. Background technique [0002] Under low temperature conditions, icing on transmission lines is a common phenomenon, which is affected by the comprehensive influence of environmental humidity, altitude, terrain conditions, wind speed, temperature and other conditions at that time. Especially in high-altitude or mountainous areas, if there is sufficient water vapor in the severe winter environment, the transmission lines will be covered with ice. Icing for a long time will gradually increase the thickness of the ice layer. When the thickness exceeds the design bearing capacity of the transmission line, it may cause the tower to fall or the wire to break. The ice layer is connected, and the flashover phenomenon occurs, which endangers the safe and stable operation of th...

AI Technical Summary

Problems solved by technology

[0004] The traditional use of the GrabCut algorithm to extract the insulators in the picture needs to roughly divide a rectangular area (foreground frame) containing all the foreground and part of the background by artificial means, and at the same time it is necessary to obtain the coordinates of the foreground frame. This method has the following problems: First, The accuracy of artificially dividing the foreground frame is insufficient, and the entire foreground may not be included in the frame, or there are too many background parts in the frame

Both of these problems will lead to a decrease in the accuracy and speed of the processing results, and even mis-segmentation. The second is that it is difficult to accurately obtain the coordinates of the foreground frame manually set, and there are errors. When the image needs to be finely processed, the impact of these errors Amplified to affect the algorithm results

The third is the efficiency of manual image processing. If there are a large number of images that need to be processed manually, it will take time and effort, and a lot of repetitive work will further reduce the accuracy of image processing personnel in dividing the foreground frame, thus aggravating the negative impact of the first and second problems.

Images