Cloud-based insulator recognition and defect detection models. To enhance the accuracy of recognizing specific targets, models deployed on servers are often designed with a large number of parameters and high computational complexity. Zheng et al. [
11] proposed a target detection model based on an improved fusion single shot multibox detector to address poor detection performance in infrared insulator images under complex backgrounds. This model improves feature extraction through a feature enhancement module and multi-scale feature maps while employing clustering algorithms to tackle the challenge posed by the large aspect ratios of insulators. Gao et al. [
12] proposed a detection network that integrates the BN-CBAM with a feature fusion module to detect defects in small insulator strings. To improve the precision of identifying insulators and detecting their defects, Zhao et al. [
13] used an improved Faster R-CNN based on the Feature Pyramid Network (FPN) to locate insulators in complex backgrounds, employing a threshold segmentation algorithm, line detection, and vertical projection for fault detection. To tackle the issue of small insulator fault regions, Zhang et al. [
14] incorporated a DC-FPN into the YOLOv3 network. Zhang et al. [
15] enhanced the YOLOv5 algorithm by integrating the squeeze-and-excitation (SE) module to solve insulator defect detection. Ou et al. [
16] improved the Faster R-CNN model by refining the feature extraction network and adding extra detection box aspect ratios for detecting various electrical equipment. Fu et al. [
17] introduced an I2D-Net that incorporates multiple novel modules to enhance feature extraction from shallow layers, including a TFFN, an RFA+ module, and a CPM to suit various scales and backgrounds. He et al. [
5] proposed an improved YOLOv8 algorithm, MFI-YOLO, for detecting multiple insulator fault types, utilizing a C2F network built with GhostNet and MSA-GhostBlock for feature extraction in complex backgrounds, and a multi-scale feature fusion structure called ResPANet to improve detection accuracy in multi-target scenarios. To detect severely occluded insulators in power lines, Luo et al. [
6] proposed an occlusion insulator detection system based on YOLOX, using an improved SPP module and the AFF-BiFPN for effective extraction of occluded and small defect information, and a coarse extraction method based on adaptive anchor frames to enhance detection performance. To address significant changes in target scale in insulators, Wang et al. [
3] proposed the multiscale channel information (MCI)-global-local attention (GLA), a plugin designed for YOLO series models, consisting of the MCI extraction module and the GLA based on context information module (GLA-CI). The MCI module extracts and utilizes multi-scale feature map information comprehensively, while the GLA-CI module captures global context information and local spatial details, enhancing learning capabilities. A comparative analysis of cloud-based insulator recognition and defect detection models is shown in
Table 1, highlighting their improvements. These methods generally involve incorporating the latest attention mechanisms into the backbone, more powerful feature extraction networks, or adaptive anchor frame algorithms in the model head to address the challenge of detecting insulators with large aspect ratios, without specifically considering their characteristics, often leading to missed or false detections.
Edge-based insulator recognition and defect detection models. In such tasks, models deployed on edge devices are typically required to have fewer parameters and lower computational complexity. Yang et al. [
18] enhanced the FPN structure of YOLOv3 to a Bidirection-Fusion FPN to improve the detection of small objects, incorporating EIoU and smooth-EIoU loss functions to speed up convergence during the training phase, thus enhancing both processing speed and recognition accuracy. To tackle the difficulty of identifying small-scale insulator defects in images, Hao et al. [
19] proposed an improved YOLOv4-based insulator defect detection model named ID-YOLO. This model uses a CSP-ResNeSt as the backbone, introducing a multiscale Bi-SimAM-FPN for more effective feature fusion. Li et al. [
7] introduced a PEDNet based on YOLOv4-tiny to enhance detection accuracy and speed in infrared images of electrical equipment. This model proposes a novel GIAM to focus the network’s attention on salient regions, integrating an FEFN in the backbone for comprehensive feature integration. To address low detection efficiency caused by large model parameters, Feng et al. [
8] developed an improved YOLOv4 model—YOLOv4++. It utilizes MobileNetv1 as the backbone and substitutes standard convolution with DSConv. Lu et al. [
4] proposed an IDD-YOLO designed for deployment on UAV hardware. This model adopts GhostNet for the backbone, incorporates a lightweight attention mechanism (LCSA) for more comprehensive feature capture, and uses PANet for feature transformation. As shown in
Table 2 these algorithms typically employ lightweight networks in the backbone to reduce parameter count and use the latest attention mechanisms in the neck to improve detection accuracy. Although general lightweight algorithms exhibit strong generalization capabilities, insulators are often located in complex terrains such as wilderness and mountainous areas. Moreover, extreme weather conditions can further increase the likelihood of missed detections and false detections of insulators and their defects [
4].
Cloud–edge collaborative insulator recognition and defect detection models. Currently, various approaches have started exploring cloud–edge collaborative strategies to tackle tasks related to insulator identification and fault detection. Song et al. [
10] introduced an intelligent cloud–edge collaborative method designed for insulator identification and fault detection within the power industrial IoT. They developed an algorithm for estimating potential insulator orientations on the drones, complemented by a defect detection method on the servers. However, images captured by UAVs tend to be large, resulting in numerous iterations needed for the orientation estimation algorithm. Furthermore, if it were possible to directly determine whether an image contains potential insulator targets, it could further reduce the drone’s energy consumption. To address the issues of significant resource consumption and high latency inherent in traditional centralized cloud computing, Wei et al. [
9] designed a lightweight SSD object recognition network on the edge. On the cloud, they employed three distinct detection algorithms and utilized a multi-model fusion algorithm to obtain the coordinates and confidence level of self-explosion regions in insulators. A comparative analysis of cloud–edge collaborative insulator recognition and defect detection models is presented in
Table 3, highlighting their improvements. However, directly applying object detection models intended for generic objects fails to adequately account for the unique large aspect ratio of insulators, which poses challenges in improving recognition accuracy.