Image Acquisition of Power Line Transmission Towers Using UAV and Deep Learning Technique for Insulators Localization and Recognition

Ohta, Hiroshi; Sato, Yuito; Mori, Takashi; Takaya, Kenta; Kroumov, Valeri

doi:10.1109/icstcc.2019.8885695

Cited by 15 publications

(7 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Whereas [30], [49] worked on porcelain and composite insulator types, Ling et al [31] explored glass insulators. As pointed out in [50], it is common to find a tower with a combination of insulator strings. Li et al [51] used a similar method to [31] for global detection and local segmentation of glass and ceramic insulators but added online hard example mining to deal with class imbalance between foreground and background.…”

Section: B Previous Work On Automated Monitoringmentioning

confidence: 99%

“…The model was further fine-tuned using a more specific data set of porcelain and composite insulators and considering scenes with vegetation, roof-tops, etc. YOLO has also been used for insulator detection [50], [54]. Han et al [55] used a ResNet50 to extract insulator features for YOLO, and considered different input scales.…”

Section: B Previous Work On Automated Monitoringmentioning

confidence: 99%

“…Authors use different proprietary data sets, which are often insufficient in VOLUME 4, 2021 size. Data augmentation has been used to artificially increase the size of small data sets [30], [31], [36], [38], [49], [50], [55] as well as to address imbalance between normal and abnormal targets, e.g., 60 defective insulators photographed from a UAV were used to generate a further 996 images using data augmentation in [36]. Images have also been sourced from the web [50].…”

Section: B Previous Work On Automated Monitoringmentioning

confidence: 99%

See 2 more Smart Citations

Aerial Image Analysis Using Deep Learning for Electrical Overhead Line Network Asset Management

Odo

McKenna²,

Flynn

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Electricity networks are critical infrastructure, delivering vital energy services. Due to the significant number, variety and distribution of electrical network overhead line assets, energy network operators spend millions annually on inspection and maintenance programmes. Currently, inspection involves acquiring and manually analysing aerial images. This is labour intensive and subjective. Along with costs associated with helicopter or drone operations, data analysis represents a significant financial burden to network operators. We propose an approach to automating assessment of the condition of electrical towers. Importantly, we train machine learning tower classifiers without using condition labels for individual components of interest. Instead, learning is supervised using only condition labels for towers in their entirety. This enables us to use a real-world industry dataset without needing costly additional human labelling of thousands of individual components. Our prototype first detects instances of components in multiple images of each tower, using Mask R-CNN or RetinaNet. It then predicts tower condition ratings using one of two approaches: (i) component instance classifiers trained using class labels transferred from towers to each of their detected component instances, or (ii) multiple instance learning classifiers based on bags of detected instances. Instance or bag class predictions are aggregated to obtain tower condition ratings. Evaluation used a dataset with representative tower images and associated condition ratings covering a range of component types, scenes, environmental conditions, and viewpoints. We report experiments investigating classification of towers based on the condition of their multiple insulator and U-bolt components. Automated visual detection and analysis of U-bolts has not been previously reported. We demonstrate that tower condition classifiers can be trained effectively without labelling the condition of individual components.

show abstract

Section: B Previous Work On Automated Monitoringmentioning

confidence: 99%

Section: B Previous Work On Automated Monitoringmentioning

confidence: 99%

Section: B Previous Work On Automated Monitoringmentioning

confidence: 99%

See 1 more Smart Citation

Aerial Image Analysis Using Deep Learning for Electrical Overhead Line Network Asset Management

Odo

McKenna²,

Flynn

et al. 2021

IEEE Access

View full text Add to dashboard Cite

show abstract

“…In [11] the SSD combined with the deep residual networks was used to discern various device defects of transmission lines. Reference [12] used YOLO v3 [13] to detect insulators and transmission towers. Reference [14] utilized YOLO v3 to simultaneously detect insulators and string drop defects in the original image.…”

Section: A Related Workmentioning

confidence: 99%

Insulator Defect Recognition Based on Global Detection and Local Segmentation

Gaohua

2020

IEEE Access

View full text Add to dashboard Cite

Locating the tiny insulator defect object with complex backgrounds in high-resolution aerial images is a challenging task. In this paper, we propose a novel method which cascades detection and segmentation networks to identify the defect from the global and local two levels: (1) The improved Faster R-CNN is carried out to capture both defects and insulators in the entire image. ResNeXt-101 is adopted as the feature extraction network so as to fully extract features, and Feature Pyramid Network (FPN) is built to enhance the ability of detecting small targets. In addition, the Online Hard Example Mining (OHEM) training strategy is applied to solve the imbalance problem of positive and negative samples. (2) All the detected insulators are extracted and fed into the improved U-Net network to futher inspect at pixel level, we utilize the pre-trained ResNeXt-50 as the encoder of U-Net, incorporate an attention module, Spatial and Channel Squeeze & Excitation Block (SCSE), into the decoding path to highlight the meaningful information. A hybrid loss which merges binary cross entropy (BCE) loss and dice coefficient loss is designed to train our network for figuring out the class imbalance issue. The missed detection can be greatly reduced with the combination of two modified network, which makes comprehensive use of the original map information and local information. On the test set of actual images, the insulator defect recognition precision and recall of the cascade network is 91.9% and 95.7%, exhibiting strong robustness and accuracy.

show abstract

“…Hence, a significant amount of research is addressing this field. Many researchers have applied computer vision techniques for power transmission towers, and insulators recognition [2][3][4][5][6][7][8]. Image processing algorithms are also heavily employed to power lines recognition and tracking [2,[9][10][11][12][13][14][15][16][17].…”

Section: Introductionmentioning

confidence: 99%

Tracking Control of Unmanned Aerial Vehicle for Power Line Inspection

Takaya¹,

Ohta²,

Shibayama³

et al. 2022

Motion Planning

Self Cite

View full text Add to dashboard Cite

This work presents some results about power transmission line tracking control and a full autonomous inspection using a quadrotor helicopter. The presented in this paper power line autonomous inspection allows detecting power line defects caused by thunderstorms, corrosion, insulator malfunctions, and same time monitoring of vegetation under the power line corridor. Traditional inspection is performed by helicopters equipped with high-resolution cameras or by direct visual examination carried out by highly skilled staff climbing over de-energized power lines. However, the visual inspection is time-expensive and costly. Moreover, due to regulatory constraints, the helicopters cannot cover narrow mountainous areas. Unmanned aerial vehicles (UAV) are an attractive alternative for power line inspection. In this work, a mathematical model for the quadrotor helicopter used in the autonomous inspection is presented. The model is successfully evaluated through simulations and flight experiments. Next, the construction of a quadrotor helicopter system and its application to power line autonomous inspection is introduced. Simulation and experimental results demonstrate the efficiency and applicability of that system. The results of this research are in the process of implementation for regular inspection of electrical transmission lines.

show abstract

Image Acquisition of Power Line Transmission Towers Using UAV and Deep Learning Technique for Insulators Localization and Recognition

Cited by 15 publications

References 9 publications

Aerial Image Analysis Using Deep Learning for Electrical Overhead Line Network Asset Management

Aerial Image Analysis Using Deep Learning for Electrical Overhead Line Network Asset Management

Insulator Defect Recognition Based on Global Detection and Local Segmentation

Tracking Control of Unmanned Aerial Vehicle for Power Line Inspection

Contact Info

Product

Resources

About