Image recognition method and system for power transmission line machine patrol

Abstract

The invention provides an image recognition method for power transmission line machine patrol. The method comprises the steps of obtaining a first image through image preprocessing; carrying out pixel-level segmentation processing on the first image through a full convolutional neural network to obtain a pixel position of an insulator, carrying out pattern corrosion expansion processing on the pixel position of the insulator, and removing noise points to obtain a connected region of the insulator; scanning the connected region, carrying out defect position prediction and type prediction through a deep convolutional neural network, carrying out non-maximum suppression processing on a prediction result, and removing an overlapping region. According to the image recognition method for power transmission line machine patrol, the image technology and the space technology are used as basic technical standards, the unmanned aerial vehicle patrol maintenance management level of the power transmission line is improved, convenient maintenance is achieved, the unmanned aerial vehicle line patrol management efficiency is improved, and the patrol cost is reduced.

Description

technical field [0001] The invention relates to the technical field of electric power equipment, in particular to an image recognition method and system for a transmission line machine patrol. Background technique [0002] With the development of long-distance power transmission technology, the terrain of power transmission corridors is becoming more and more complex, the types of poles and towers are becoming more and more diverse, and the natural environment is getting worse and worse. Coupled with the influence of external factors such as material aging, power grids are prone to accidents. If the problem of equipment defects is not dealt with in time, it is likely to seriously threaten the stability and security of the power system. [0003] The traditional manual inspection operation method has high cost and low efficiency, which has been difficult to meet the inspection requirements of transmission lines. Traditional UAV power inspection uses visible light to obtain im...

AI Technical Summary

Problems solved by technology

[0002] With the development of long-distance power transmission technology, the terrain of power transmission corridors is becoming more and more complex, the types of poles and towers are becoming more and more diverse, and the natural environment is getting worse and worse. Coupled with the influence of external factors such as material aging, power grids are prone to accidents. If the problem of equipment defects is not dealt with in time, it is likely to seriously threaten the stability and security of the power system

[0003] The traditional manual inspection method has high cost and low efficiency, and it has been difficult to meet the inspection requirements of transmission lines.

Traditional UAV power inspection uses visible light to obtain image data of transmission lines for fault analysis, but due to the lack of software for visible light space technology analysis, when using UAV to detect camphor, it needs to be equipped with laser radar for inspection, resulting in The inspection cost is too high and the inspection time is too long, which is not conducive to quickly detecting the distance between the tree and the line

Traditional UAVs are equipped with laser mounts for tree camphor inspections. The cost of inspections is too high and the inspection time is too long, which is not conducive to quickly detecting the distance between trees and lines. Inconvenience seriously affects the popularization and inspection quality of multi-rotor drones

Images