Power transmission and transformation line insulator aerial image fault detection method based on improved YOLOv3

Abstract

The invention discloses a power transmission and transformation line insulator aerial image fault detection method based on improved YOLOv3. The method comprises the following steps of S1, establishing an insulator data set and an insulator fault data set; s2, performing data enhancement by using modes of rotation, overturning, contrast enhancement and the like; s3, taking YOLOv3 as a basic framework, carrying out model modification pruning operation, and constructing an insulator detection model; s4, establishing a standard YOLOv3-tiny network; s5, carrying out model training on the modifiedYOLOv3 network and the modified YOLOv3-tiny by using the insulator data set and the insulator fault data set; and S6, connecting the two trained models to realize end-to-end rapid detection of the insulator fault. According to the detection method, the insulator in the complex background image can be quickly and accurately identified, the fault area can be positioned and detected, the inspection efficiency is improved, the working intensity of inspection personnel is reduced, and a guarantee is provided for normal operation of a power system.

Description

technical field [0001] The invention belongs to the field of deep learning computer vision and the field of electric defect identification, and in particular relates to an aerial image fault detection method for power transmission and transformation line insulators based on improved YOLOv3. Background technique [0002] There are a large number of insulators in power transmission and transformation lines, which play an important role in electrical insulation and mechanical connection. Because the power transmission and transformation lines span various complex natural and geographical environments and are exposed to wind, rain and sun for a long time, they are prone to various failures including self-explosion and flashover. Once an insulator fails, it will seriously affect the normal and safe transmission of power, thus threatening the normal operation of the power system. Therefore, fast and efficient fault detection of insulators is one of the important procedures for the...

AI Technical Summary

Problems solved by technology

However, the traditional target detection algorithm mainly has three defects. First, the designed features are low-level features, which are not expressive enough for the target; The robustness is not very good; finally, for the current mass inspection pictures, the detection speed and detection effect cannot meet the actual industrial application requirements

[0004]Deep learning has developed rapidly since 2012 and has achieved good results in the field of computer vision. However, there are few applications of deep learning for power inspection at present. , and cannot meet the real-time and accuracy requirements of power transmission and transformation line inspection. Therefore, this invention mainly studies the application of YOLOv3 model to the fault detection of insulator aerial inspection images to achieve end-to-end efficient and accurate detection

Images