Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 17, 2025MDPIOpen Access

Research on Void and Defect Detection in Ground-Penetrating Radar Images Using Deep Learning Techniques

View Full Paper
Ask AI
Bookmark
Share

Authors

KHKeng-Tsang HsuYWYinghui Wang

Discussion

Loading...

Member takes

Overview

This research demonstrates the effectiveness of deep learning in detecting embankment cavities using GPR images, highlighting accuracy improvements.

Key Points

  • The YOLOv10 model identifies cavities in ground-penetrating radar images with high accuracy rates of nearly 90% and 97%.
  • Key strategies include dataset expansion through data augmentation and fine-tuning hyperparameters for optimal performance.
  • This study showcases how deep learning enhances detection efficiency and accuracy, contributing to better embankment safety inspections.
  • Ground-penetrating radar's traditional image interpretation challenges are addressed through advanced automated techniques.

Cite This Study

Hsu et al. (2025) studied this question.

synapsesocial.com/papers/68d4604031b076d99fa5f421https://doi.org/10.3390/proceedings2025129031
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1An Improved YOLOv11-Based Detection Method for Hidden Void and Loose Defects in Urban Road Ground-Penetrating Radar Images2026
  2. 2Field Validation of Deep-Learning-Based Ground Penetrating Radar Image Analysis for Advancing Subsurface Distress Detection2024 · 1 citations
  3. 3Three‐dimensional reconstruction of loose defects in semi‐rigid base layers using enhanced deep learning and point cloud from GPR images2025
  4. 4Three‐dimensional reconstruction of loose defects in semi‐rigid base layers using enhanced deep learning and point cloud from GPR images2025 · 2 citations
  5. 5A Simple Augmentation Method Using Cutout for Ground Penetrating Radar Image in Deep Learning2024