F-RCNN-based defective cable detection method

Abstract

The invention belongs to the field of cable defect identification, and particularly relates to a CNN-based defective cable detection method. The method is mainly characterized by including the following steps: 1) combining a fast regional convolutional neural network (F-RCNN) method with an image shooting technology; 2) extracting features of images by using a convolutional neural network (CNN); 3) randomly combining the features extracted by the convolutional neural network (CNN) by using a regional convolutional neural network (RCNN) to avoid the possibility of each feature from being ignored; and 4) fast calculating any combination of the previous step by using the fast regional convolutional neural network (F-RCNN). The beneficial effects of the invention are that: a lot of time and cost can be saved, the accuracy of identification can be improved, and the position of defective cables can be quickly determined.

Description

technical field [0001] The invention belongs to the field of cable defect identification, and in particular relates to an F-RCNN-based defect cable detection method. Background technique [0002] In the early 1990s, my country carried out a large number of urban network transformation work, and a large number of cables began to be applied to urban power grids. However, due to the influence of the manufacturing process and long-term operating conditions, power cables often suffer from moisture, overheating, extrusion, excessive bending, etc., resulting in water trees, deformation of the insulation medium, loose copper shielding, etc. Local defects, if the local defects of the power cable are not dealt with, the local defects of the power cable will develop rapidly under the action of the strong electric field, and eventually lead to the failure of the power cable insulation, thus bringing a lot of work to the maintenance and replacement of the power cable. Result in a lot of...

AI Technical Summary

Problems solved by technology

The existing problems are: first, the number of photos taken by the UAV is large, and storage is difficult; second, the UAV can only take a certain discrete point, and cannot record the entire profile of the cable; third, the number of photos is large, When screening, it is time-consuming and labor-intensive, and it is easy to make mistakes; Fourth, sometimes it is easily affected by the external environment, such as wind and sand, which leads to blurred photos

Images