{"ID":2894295,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2507.11081","arxiv_id":"2507.11081","title":"Automatic Road Subsurface Distress Recognition from Ground Penetrating Radar Images using Deep Learning-based Cross-verification","abstract":"Ground penetrating radar (GPR) has become a rapid and non-destructive solution for road subsurface distress (RSD) detection. However, recognizing RSD from GPR images is labor-intensive and heavily relies on the expertise of inspectors. Deep learning-based automatic RSD recognition, though ameliorating the burden of data processing, suffers from insufficient capability to recognize defects. In this study, a novel cross-verification strategy was proposed to fully exploit the complementary abilities of region proposal networks in object recognition from different views of GPR images. Following this strategy, three YOLO-based models were used to detect the RSD (voids and loose structures) and manholes. Each model was trained with a specific view of 3D GPR dataset, which contains rigorously validated 2134 samples of diverse types obtained through field scanning. The cross-verification strategy achieves outstanding accuracy with a recall of over 98.6% in the tests using real field-scanning data. Field tests also show that deep learning-based automatic RSD recognition can reduce the human labor of inspection by around 90%.","short_abstract":"Ground penetrating radar (GPR) has become a rapid and non-destructive solution for road subsurface distress (RSD) detection. However, recognizing RSD from GPR images is labor-intensive and heavily relies on the expertise of inspectors. Deep learning-based automatic RSD recognition, though ameliorating the burden of dat...","url_abs":"https://arxiv.org/abs/2507.11081","url_pdf":"https://arxiv.org/pdf/2507.11081v3","authors":"[\"Chang Peng\",\"Bao Yang\",\"Meiqi Li\",\"Ge Zhang\",\"Hui Sun\",\"Zhenyu Jiang\"]","published":"2025-07-15T08:23:21Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.AI\"]","methods":"[]","has_code":false}