Damage detection on steel-reinforced concrete produced by corrosion via YOLOv3: A detailed guide

Damage assessment applied to reinforced concrete elements is one of the main activities infrastructure maintenance tasks. Among these elements, problem corrosion in particularly critical and requires careful consideration. Annually, governments invest a large amount economic resources this activity. However, most methodologies for damage rely on visual inspection, which may be subjectively interpreted, producing inconsistent results requiring considerable time resources. This study evaluates performance real-time object detection using You Only Look Once, version 3, detecting structures. The architecture YOLOv3 based complex, but efficient, convolutional neural network fed by dataset proposed labeled authors. Two training stages were established improve model precision, transfer learning with medium- high-resolution images. test show satisfactory concrete-corrosion through validation photographs videos demonstrating capabilities explainable artificial intelligence its applications civil engineering.