Article: Change detection framework for power facilities in disaster scenarios Journal: International Journal of Power and Energy Conversion (IJPEC) 2026 Vol.17 No.5 pp.1 - 20 Abstract: In order to regularly monitor geological surface changes around the power transmission line corridors and to swiftly and effectively respond to the negative impacts of natural disasters (specifically landslides) on electrical facilities, thereby enhancing the disaster resistance and recovery speed of the power system, while ensuring the stability of the tower foundations, a change detection model oriented towards disaster scenarios is proposed. Firstly, a twin-network-based change detection framework is designed, which utilises pre-processed remote sensing images before and after the disaster to extract multi-scale visual features using a basic visual model and a designed change detection module, respectively. Secondly, to filter effective feature information and integrate the features from both modules, an attention mechanism-based alignment module is proposed to fuse general features and domain-specific features, guiding the domain-specific features with general features for efficient feature extraction, thus ensuring the robustness and accuracy of the proposed model. Finally, experiments are conducted on real disaster datasets, and the proposed method is compared with excellent algorithms proposed in recent years to verify the accuracy of the method. Inderscience Publishers - linking academia, business and industry through research

Title: Change detection framework for power facilities in disaster scenarios

Authors: Jiangyi Wu; He Tang; Guangdu Cen; Ke Wang

Addresses: Foshan Power Supply Bureau, Guangdong Power Grid Co., Ltd., Foshan 528000, China ' Foshan Power Supply Bureau, Guangdong Power Grid Co., Ltd., Foshan 528000, China ' China Southern Power Grid Technology Co., Ltd., Guangzhou 510000, China; Guangdong Engineering Technology Research Center of Special Robots for Special Industries, No. 1, Fenjiang South Road, Chancheng District, Foshan City, Guangdong Province, 510000, China ' China Southern Power Grid Technology Co., Ltd., Guangzhou 510000, China; Guangdong Engineering Technology Research Center of Special Robots for Special Industries, No. 1, Fenjiang South Road, Chancheng District, Foshan City, Guangdong Province, 510000, China

Abstract: In order to regularly monitor geological surface changes around the power transmission line corridors and to swiftly and effectively respond to the negative impacts of natural disasters (specifically landslides) on electrical facilities, thereby enhancing the disaster resistance and recovery speed of the power system, while ensuring the stability of the tower foundations, a change detection model oriented towards disaster scenarios is proposed. Firstly, a twin-network-based change detection framework is designed, which utilises pre-processed remote sensing images before and after the disaster to extract multi-scale visual features using a basic visual model and a designed change detection module, respectively. Secondly, to filter effective feature information and integrate the features from both modules, an attention mechanism-based alignment module is proposed to fuse general features and domain-specific features, guiding the domain-specific features with general features for efficient feature extraction, thus ensuring the robustness and accuracy of the proposed model. Finally, experiments are conducted on real disaster datasets, and the proposed method is compared with excellent algorithms proposed in recent years to verify the accuracy of the method.

Keywords: change detection; CD; deep learning; disaster monitoring; remote sensing; RS; vision foundation models; VFMs; Siamese neural network.

DOI: 10.1504/IJPEC.2026.154216

International Journal of Power and Energy Conversion, 2026 Vol.17 No.5, pp.1 - 20

Received: 26 Aug 2025
Accepted: 01 May 2026
Published online: 16 Jun 2026 *

Title: Change detection framework for power facilities in disaster scenarios

Keep up-to-date