An Improved Anomaly Detection Model for Automated Inspection of Power Line Insulators

TL;DR

This paper introduces a two-stage anomaly detection approach combining object detection and explainable one-class classification to improve incipient fault detection in power line insulators, especially in data-scarce scenarios.

Contribution

It proposes a novel anomaly detection method with a modified loss function that enhances fault detection accuracy and interpretability, reducing reliance on extensive faulty insulator images.

Findings

The model performs well with as few as five anomalies in training.

The new loss function outperforms existing methods on MVTec-AD dataset.

Effective in both data-abundant and data-scarce scenarios.

Abstract

Inspection of insulators is important to ensure reliable operation of the power system. Deep learning is being increasingly exploited to automate the inspection process by leveraging object detection models to analyse aerial images captured by drones. A purely object detection-based approach, however, suffers from class imbalance-induced poor performance, which can be accentuated for infrequent and hard-to-detect incipient faults. This article proposes the use of anomaly detection along with object detection in a two-stage approach for incipient fault detection in a data-efficient manner. An explainable convolutional one-class classifier is adopted for anomaly detection. The one-class formulation reduces the reliance on plentifully available images of faulty insulators, while the explainability of the model is expected to promote adoption by the industry. A modified loss function is developed that addresses computational and interpretability issues with the existing model, also allowing for the integration of other losses. The superiority of the novel loss function is demonstrated with MVTec-AD dataset. The models are trained for insulator inspection with two datasets -- representing data-abundant and data-scarce scenarios -- in unsupervised and semi-supervised settings. The results suggest that including as few as five real anomalies in the training dataset significantly improves the model's performance and enables reliable detection of rarely occurring incipient faults in insulators.