High-voltage line insulator defect detection method and system based on deep belief network

Abstract

The invention provides a high-voltage line insulator defect detection method and system based on a deep belief network. The method includes: obtaining an original image containing the detected high-voltage line insulator; preprocessing the original image, and using the fuzzy C-means clustering method to process the processed Segment the image to obtain the high-voltage line insulator area; use the connected area identification method to trace the boundary of the high-voltage line insulator area to obtain the outline of the high-voltage line insulator, and then obtain the image of the high-voltage line insulator; extract the image feature vector of the high-voltage line insulator image; use the image feature vector as the depth The input of the belief network, and then detect the defect situation of the high voltage line insulator through the deep belief network. It can completely extract insulator images from complex aerial images, effectively detect defective high-voltage line insulators on high-voltage lines, and has the advantages of low detection cost and high detection accuracy, so it has a good application prospect.

Description

technical field [0001] The invention belongs to the technical field of high-voltage line insulator defect detection, and in particular relates to a high-voltage line insulator defect detection method and system based on a deep belief network. Background technique [0002] For overhead high-voltage lines (including high-voltage line insulators, wires, etc.), due to long-term exposure to the external natural environment, there are both man-made damage factors and natural damage or line aging factors. For example, they need to withstand pollution, lightning strikes, wind and snow, floods and Bird damage and other external factors, therefore, damage to high-voltage line insulators or wires will occur. For the high-voltage line insulator, it is one of the important parts of the overhead high-voltage line. It is used to prevent the live parts of the high-voltage line from forming a grounding channel. Accidents directly affect the safe operation of the power grid and the stability...

AI Technical Summary

Problems solved by technology

[0006] The patent with the application number "201310061850.8" and the title of the invention "A Method and Device for Toughened Glass Insulator Defect Detection" discloses a method using concentric circle scanning, which suppresses a large number of irrelevant backgrounds and leaves only the image of the defect part, which is beneficial to support The vector machine is used for defect detection. This method uses an adaptive threshold method for image segmentation, which has limited image segmentation effects, and ultimately leads to limited defect detection of high-voltage line insulators.

[0007] It can be seen that the various detection methods for high-voltage line insulator defects disclosed in the prior art either have the problem of high detection cost or the problem of limited detection accuracy.

Images