Deep-learning-based method for identifying defect of power transmission equipment automatically
According to the invention, a body photo, obtained by routing inspection by an unmanned aerial vehicle, of a transmission tower is analyzed automatically by using a deep learning technology. On the basis of a Faster-Rcnn algorithm, power transmission equipment needing detection is identified from th...
Saved in:
Main Authors | , , , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
06.07.2018
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | According to the invention, a body photo, obtained by routing inspection by an unmanned aerial vehicle, of a transmission tower is analyzed automatically by using a deep learning technology. On the basis of a Faster-Rcnn algorithm, power transmission equipment needing detection is identified from the photo obtained by routing inspection by the unmanned aerial vehicle; the power transmission equipment is sent to a defect classifier for the equipment to determine whether the equipment has a defect and which type the defect is; if the equipment has a defect, the location of the defect equipment as well as the defect type is marked in the photo automatically; and then a defect report is generated finally and maintenance information is provided for the maintenance staff. According to the invention, the analysis process has advantages of high accuracy, fast processing speed and high reliability; and the automatic analysis of the inspection photo of the transmission line is realized without manual participation.
本发明 |
---|---|
Bibliography: | Application Number: CN201711476547 |