Power grid image line anomaly detection method based on deep learning feature extraction

By employing deep learning feature extraction and adaptive threshold segmentation algorithms, the problems of low efficiency and insufficient accuracy in power grid line anomaly detection have been solved, enabling real-time early warning and operation and maintenance scheduling of power grid line anomalies, and improving the accuracy and reliability of detection.

CN121921310BActive Publication Date: 2026-07-03STATE GRID SICHUAN ELECTRIC POWER CORP ELECTRIC POWER RES INST
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
STATE GRID SICHUAN ELECTRIC POWER CORP ELECTRIC POWER RES INST
Filing Date
2026-03-24
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

Existing methods for detecting power grid line anomalies rely on manual inspections and simple sensors, which are inefficient and have limited detection accuracy. They are difficult to accurately identify line anomalies in complex environments, especially minor damage to insulators and hanging foreign objects.

Method used

A deep learning-based feature extraction method is adopted. Visible light image data of power grid lines are acquired through an image acquisition device. After preprocessing, the data is input into a pre-trained convolutional neural network model to extract multi-level features. Low-level texture features and high-level semantic features are fused. Anomaly region localization is performed by combining adaptive threshold segmentation and optimization algorithms to generate a formatted anomaly report.

Benefits of technology

It enables real-time early warning and operation and maintenance scheduling of power grid line anomalies, improves the accuracy and reliability of detection, reduces false detections and missed detections, and enhances the efficiency and reliability of power grid operation and maintenance management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121921310B_ABST
    Figure CN121921310B_ABST
Patent Text Reader

Abstract

The application relates to a power grid image line anomaly detection method and system based on deep learning feature extraction, which comprises the following steps: acquiring visible light image data of a line through an image acquisition device arranged in a power grid line area and preprocessing the visible light image data to obtain a standardized image data set; inputting the standardized image data set into a pre-trained convolutional neural network model to extract multi-level feature representations and generate a feature set; performing feature vector fusion processing and multi-class classification calculation, and outputting an anomaly probability distribution and an anomaly type identifier; positioning and correcting an anomaly area according to the anomaly probability distribution and the anomaly type identifier to generate a final anomaly detection result; and generating a formatted anomaly report based on the final anomaly detection result and sending the formatted anomaly report to a power grid monitoring center. The scheme can improve the accuracy and reliability of power grid line anomaly detection, realizes real-time early warning and operation and maintenance scheduling of line anomalies.
Need to check novelty before this filing date? Find Prior Art