Vision-based power transmission line tower online identification and inclination detection method

Abstract

The invention discloses a vision-based power transmission line tower online identification and inclination detection method, and belongs to the field of image processing and power transmission line detection. The implementation method comprises the following steps: carrying out angle correction on an aerial photo of the unmanned aerial vehicle; preprocessing the acquired image and extracting linesegment features by using an LSD algorithm; filtering the obtained line segment features of the picture to obtain line segments with the filtered inclination smaller than a preset threshold value, andfusing the filtered interrupted line segments meeting a preset line segment fusion standard; searching an outermost tower foot line segment of the tower according to the outermost tower foot characteristics, and identifying and positioning a tower area, wherein the outermost tower foot characteristics refer to that the outermost tower foot line segment of the tower meets two corresponding constraint conditions; calculating included angles between the central lines of the tower foot line segments on the outermost sides of the two towers and the vertical direction, the included angles being thecalculated inclinations of the towers in the direction, and respectively calculating the inclinations of the towers in four directions around the towers to realize inclination detection of the towers.

Description

technical field [0001] The invention belongs to the field of image processing and transmission line detection, in particular to an LSD (Line Segment Detector)-based online recognition and tilt detection method for towers. Background technique [0002] In the maintenance work of power grid transmission lines, the fault analysis of transmission line towers is a very important direction. As the main support of the transmission line, the tower directly affects the safe operation of the power grid. [0003] The method of manual inspection has high time cost, many human resources required, low efficiency and high risk, which can no longer meet the new inspection requirements. In recent years, with the advancement of drone technology, there are more and more occasions where drones are used instead of manual labor, and drone inspections have become an emerging inspection method. UAV inspection has the advantages of low cost and simple operation, and is suitable for the maintenance...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the following technical problems existing in the prior art: (1) The algorithm for target detection of towers using deep learning requires a large number of effective data sets, high cost, and poor real-time performance; (2) It is difficult to distinguish the degree of inclination of towers, and manual measurement is time-consuming laborious and inefficient

Images