Citation: Zhang, Z.; Chen, H.;
Huang, S. Detection of Missing
Insulator Caps Based on Machine
Learning and Morphological
Detection. Sensors 2023, 23, 1557.
https://doi.org/10.3390/s23031557
Academic Editor: Benoit Vozel
Received: 1 January 2023
Revised: 20 January 2023
Accepted: 25 January 2023
Published: 31 January 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
Detection of Missing Insulator Caps Based on Machine
Learning and Morphological Detection
Zhaoyun Zhang * , Hefan Chen and Shihong Huang
Electronic Engineering and Intelligence College, Dongguan University of Technology, Dongguan 523000, China
* Correspondence: zhangzy@dgut.edu.cn; Tel.: +86-189-2749-1998
Abstract:
Missing insulator caps are the key focus of transmission line inspection work. Insula-
tors with a missing cap will experience decreased insulation and mechanical strength and cause
transmission line safety accidents. As missing insulator caps often occur in glass and porcelain
insulators, this paper proposes a detection method for missing insulator caps in these materials. First,
according to the grayscale and color characteristics of these insulators, similar characteristic regions
of the insulators are extracted from inspection images, and candidate boxes are generated based on
these characteristic regions. Second, the images captured by these boxes are input into the classifier
composed of SVM (Support Vector Machine) to identify and locate the insulators. The accuracy,
recall and average accuracy of the classifier are all higher than 90%. Finally, this paper proposes
a processing method based on the insulator morphology to determine whether an insulator cap is
missing. The proposed method can also detect the number of remaining insulators, which can help
power supply enterprises to evaluate the degree of insulator damage.
Keywords:
SVM; missing insulator slices; small-scale dataset; object region detection; machine
learning; morphological detection
1. Introduction
The functions of insulators, which play an important role in high-voltage transmission
lines, are to support power lines and provide electrical insulation [1–3]. Missing insulator
caps are usually found on glass and porcelain insulators, which are mainly caused by
tension on insulators from power lines and towers, and erosion of insulators by wind, acid
rain and other weather elements. The insulators of these materials age due to accumulated
weather erosion. Under the action of power line tension, the insulator cap is damaged
and falls off. In addition, glass insulators can eliminate a flashover cap, which is called
“zero-value self-explosion”. Thus, the insulator cap will be missing more frequently in
glass insulators. A missing insulator cap decreases the mechanical and electrical properties
of the whole insulator string, which will threaten power transmission system operations.
Therefore, the detection of missing insulator caps is one of the most important topics in
transmission line detection.
Computer vision, which is used to detect transmission line images collected by UAV
(Unmanned Aerial Vehicles), is the most popular transmission line detection method [
4
–
6
].
The main reason for its popularity in high-voltage transmission line detection is that its
efficiency and safety are higher than those of manual inspection and crewed helicopter de-
tection. Therefore, high-voltage transmission line detection technology based on computer
vision has great research significance and practical value [7–9].
The method of UAV transmission line detection is shown in Figure 1. It can be divided
into two types: detection images in workstation and detection images in UAV [
10
–
12
].
Transmission line images can be quickly and accurately detected by the detection method
of transmission equipment deployed in the workstation. However, this method not only
requires the UAV to transmit a large amount of data to the workstation, but struggles to
Sensors 2023, 23, 1557. https://doi.org/10.3390/s23031557 https://www.mdpi.com/journal/sensors
