AI-based Damage Detection for Power Grids: publication in MDPI Remote Sensing
GAP is pleased to share its latest technical paper published in the journal MDPI Remote Sensing:
“Applying MLP and SVM Models to Detect Potential Damages on High-Voltage Power Transmission Towers and Lines Using Multi-Temporal SAR Images”.
The study, conducted by Geophysical Applications Processing srl in collaboration with the Italian Space Agency, Planetek Italia and the Polytechnic University of Bari,
explores the use of Synthetic Aperture Radar (SAR) data for infrastructure monitoring.
The research compares Multi-Layer Perceptron (MLP) and Support Vector Machine (SVM) models on COSMO-SkyMed images.
To address the limited availability of real damage samples, a physics-guided SAR simulator was developed to supplement the training datasets.
In the validated test set, the MLP model demonstrated high reliability in identifying structural changes, with an inference latency compatible with future on-board satellite implementation.
This work highlights the potential of satellite remote sensing to support traditional inspection methods, particularly for rapid assessment following adverse weather events.
Title:
“Applying
MLP and SVM Models to Detect Potential Damage to High-Voltage Power Transmission Towers and Lines Using Multi-Temporal SAR Images”
Affiliations
:
Geophysical Applications Processing (GAP) srl, Bari, Italy
; Italian
Space Agency; Planetek Italia
; Polytechnic University of Bari
Available
resources
Full article (Open Access, MDPI Remote Sensing)
#RemoteSensing #SAR #InfrastructureMonitoring #AI #Engineering #GAPsrl #COSMOSkyMed
