Insulator visual non-conformity detection in overhead power distribution lines using deep learning

Prates, Ricardo M.; Cruz, Ricardo; Marotta, André; Ramos, Renato Moran; Filho, Eduardo F. Simas; Cardoso, Jaime S.

doi:10.1016/j.compeleceng.2019.08.001

Cited by 52 publications

(21 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The proposed method presents a higher F1-score on insulator defect detection of overhead power line distribution systems. In [ 41 ], by using multi-task learning, an F1-score of 0.75 is achieved on insulator defect detection in overhead power lines, while our proposed methodology achieved an F1-score of 0.77 on the Prototype-C test dataset and an F1-score of 0.81 on the Prototype-S test dataset. Both of these datasets have both defective pin and suspension disc insulators.…”

Section: Experimentation and Resultsmentioning

confidence: 99%

Autonomous Vision-Based Primary Distribution Systems Porcelain Insulators Inspection Using UAVs

Rahman

Zhang

Ahmad

et al. 2021

Sensors

View full text Add to dashboard Cite

The early detection of damaged (partially broken) outdoor insulators in primary distribution systems is of paramount importance for continuous electricity supply and public safety. Unmanned aerial vehicles (UAVs) present a safer, autonomous, and efficient way to examine the power system components without closing the power distribution system. In this work, a novel dataset is designed by capturing real images using UAVs and manually generated images collected to overcome the data insufficiency problem. A deep Laplacian pyramid-based super-resolution network is implemented to reconstruct high-resolution training images. To improve the visibility of low-light images, a low-light image enhancement technique is used for the robust exposure correction of the training images. A different fine-tuning strategy is implemented for fine-tuning the object detection model to increase detection accuracy for the specific faulty insulators. Several flight path strategies are proposed to overcome the shuttering effect of insulators, along with providing a less complex and time- and energy-efficient approach for capturing a video stream of the power system components. The performance of different object detection models is presented for selecting the most suitable one for fine-tuning on the specific faulty insulator dataset. For the detection of damaged insulators, our proposed method achieved an F1-score of 0.81 and 0.77 on two different datasets and presents a simple and more efficient flight strategy. Our approach is based on real aerial inspection of in-service porcelain insulators by extensive evaluation of several video sequences showing robust fault recognition and diagnostic capabilities. Our approach is demonstrated on data acquired by a drone in Swat, Pakistan.

show abstract

Section: Experimentation and Resultsmentioning

confidence: 99%

Autonomous Vision-Based Primary Distribution Systems Porcelain Insulators Inspection Using UAVs

Rahman

Zhang

Ahmad

et al. 2021

Sensors

View full text Add to dashboard Cite

show abstract

“…The proposed method presents a higher F1-score on insulator defect detection of overhead power line distribution systems. In literature [41] by using multi-task learning an F1-score of 0.75 is achieved on insulator defect detection in over-head power lines while our proposed methodology achieved an F1-score of 0.77 on the Prototype-C test dataset and F1-score of 0.81 on the Prototype-S test dataset. Both of these datasets have both defective pin and suspension disc insulators.…”

Section: Faulty Insulator Detection Using Proposed Methodology With Dmentioning

confidence: 89%

Autonomous Vision-Based Primary Distribution Systems Porcelain Insulators Inspection Using UAVs

Rahman¹,

Zhang²,

Ahmad³

et al. 2020

Preprint

View full text Add to dashboard Cite

The primary distribution systems are comprised of power lines delivering power to utility feeders from substations. The inspection and maintenance of damaged and broken power system insulators are of paramount importance for continuous power supply and public safety. hence, to identify any faults and defects in advance a periodic inspection of power line insulators and other components be ensured beforehand. To automate the process and reduce operational cost and risk Unmanned Aerial Vehicles (UAVs) are being extensively utilized. As they present a safer and efficient way to examine the power system insulators and their components without closing the power distribution system ensuring continuous supply to the end-users. To achieve these objectives in this work a novel dataset is designed by capturing real images using UAVs and manually generated images collected to overcome the data insufficiency problem. Deep Laplacian pyramids based super-resolution network is implemented to reconstruct high-resolution training images. To improve the visibility of low light images a low light image enhancement technique is used for the robust exposure correction of the training images. Using computer vision-based object detection techniques to identify faults and classify them according to classes they belong. A different fine-tuning strategy is implemented for fine-tuning the object detection model to increase detection accuracy for the specific faulty insulators. To improve the faults detection several flight path strategies are proposed for efficient inspection. Such strategies overcome the shuttering effect of insulators along with providing a less complex, time, and energy-efficient approach for capturing video stream of the power system components. Performance of different object detection models is presented for selecting the suitable one for fine-tuning on the specific faulty insulator dataset. Our proposed approach gives a less complex and more efficient flight strategy along with better results. For defect detection, our proposed method achieved an F1-score of 0.81 and 0.77 on two different datasets. Our approach is based on real aerial inspection of in-service porcelain insulators by extensive evaluation of several video sequences showing robust faults recognition and diagnostic capabilities. Our approach is demonstrated on data acquired by a drone in Swat Pakistan.

show abstract

“…As shown in Equations (13) and (14), L cls (t,p) is the category confidence loss function. When calculating category confidence errors, both positive and negative samples must be calculated.…”

Section: Loss Function and Optimizationmentioning

confidence: 99%

“…In this method, an SR-CNN (super-resolution convolutional neural network) [12] is used to reconstruct a blurred image with high resolution, and then the YOLOv3 (you only look once) [13] network is used to recognize the reconstructed image. For overhead distribution line insulator detection, Prates et al [14] proposed a deep learning-based method, which uses the VGG [15] network to classify the insulator images. However, this method does not detect the position of insulators.…”

Section: Introductionmentioning

confidence: 99%

ISSD: Improved SSD for Insulator and Spacer Online Detection Based on UAV System

Shuang

et al. 2020

Sensors

View full text Add to dashboard Cite

In power inspection tasks, the insulator and spacer are important inspection objects. UAV (unmanned aerial vehicle) power inspection is becoming more and more popular. However, due to the limited computing resources carried by a UAV, a lighter model with small model size, high detection accuracy, and fast detection speed is needed to achieve online detection. In order to realize the online detection of power inspection objects, we propose an improved SSD (single shot multibox detector) insulator and spacer detection algorithm using the power inspection images collected by a UAV. In the proposed algorithm, the lightweight network MnasNet is used as the feature extraction network to generate feature maps. Then, two multiscale feature fusion methods are used to fuse multiple feature maps. Lastly, a power inspection object dataset containing insulators and spacers based on aerial images is built, and the performance of the proposed algorithm is tested on real aerial images and videos. Experimental results show that the proposed algorithm can efficiently detect insulators and spacers. Compared with existing algorithms, the proposed algorithm has the advantages of small model size and fast detection speed. The detection accuracy can achieve 93.8%. The detection time of a single image on TX2 (NVIDIA Jetson TX2) is 154 ms and the capture rate on TX2 is 8.27 fps, which allows realizing online detection.

show abstract

Insulator visual non-conformity detection in overhead power distribution lines using deep learning

Cited by 52 publications

References 15 publications

Autonomous Vision-Based Primary Distribution Systems Porcelain Insulators Inspection Using UAVs

Autonomous Vision-Based Primary Distribution Systems Porcelain Insulators Inspection Using UAVs

Autonomous Vision-Based Primary Distribution Systems Porcelain Insulators Inspection Using UAVs

ISSD: Improved SSD for Insulator and Spacer Online Detection Based on UAV System

Contact Info

Product

Resources

About