Early and timely detection of surface damages is important for maintaining the functionality, reliability, and safety of concrete bridges. Recent advancement in convolution neural network has enabled the development of deep learning‐based visual inspection techniques for detecting multiple structural damages. However, most deep learning‐based techniques are built on two‐stage, proposal‐driven detectors using less complex image data, which could be restricted for practical applications and possible integration within intelligent autonomous inspection systems. In this study, a faster, simpler single‐stage detector is proposed based on a real‐time object detection technique, You Only Look Once (YOLOv3), for detecting multiple concrete bridge damages. A field inspection images dataset labeled with four types of concrete damages (crack, pop‐out, spalling, and exposed rebar) is used for training and testing of YOLOv3. To enhance the detection accuracy, the original YOLOv3 is further improved by introducing a novel transfer learning method with fully pretrained weights from a geometrically similar dataset. Batch renormalization and focal loss are also incorporated to increase the accuracy. Testing results show that the improved YOLOv3 has a detection accuracy of up to 80% and 47% at the Intersection‐over‐Union (IoU) metrics of 0.5 and 0.75, respectively. It outperforms the original YOLOv3 and the two‐stage detector Faster Region‐based Convolutional Neural Network (Faster R‐CNN) with ResNet‐101, especially for the IoU metric of 0.75.

中文翻译：

---

使用单级检测器检测混凝土桥梁表面损伤

及早发现表面损伤对于保持混凝土桥梁的功能，可靠性和安全性很重要。卷积神经网络的最新进展使基于深度学习的视觉检测技术得以发展，该技术可检测多种结构损伤。但是，大多数基于深度学习的技术都是基于使用较少复杂图像数据的两阶段，建议驱动的检测器构建的，这可能会受到实际应用的限制，并且可能无法集成到智能自主检查系统中。在这项研究中，提出了一种基于实时对象检测技术“仅一次查看”（YOLOv3）的更快，更简单的单级检测器，用于检测多个混凝土桥梁损伤。现场检查图像数据集，标有四种类型的混凝土破坏（裂缝，弹跳，剥落和裸露的钢筋）用于YOLOv3的培训和测试。为了提高检测精度，通过从几何相似的数据集中引入具有完全预训练权重的新颖转移学习方法，进一步改善了原始YOLOv3。批处理重归一化和焦点损失也被纳入以提高准确性。测试结果表明，改进的YOLOv3在联合交叉口（IoU）度量分别为0.5和0.75时，分别具有高达80％和47％的检测精度。它具有ResNet-101的性能优于原始的YOLOv3和两级检测器基于Faster Region-based的卷积神经网络（Faster R-CNN），尤其是对于0.75的IoU度量。原始的YOLOv3通过引入一种新颖的转移学习方法得到了进一步的改进，该方法具有来自几何相似数据集的完全预训练的权重。批处理重归一化和焦点损失也被纳入以提高准确性。测试结果表明，改进的YOLOv3在联合交叉口（IoU）度量分别为0.5和0.75时，分别具有高达80％和47％的检测精度。它具有ResNet-101的性能优于原始的YOLOv3和两级检测器基于Faster Region-based的卷积神经网络（Faster R-CNN），尤其是对于0.75的IoU度量。最初的YOLOv3通过引入一种新的转移学习方法得到了进一步的改进，该方法具有来自几何相似数据集的完全经过预训练的权重。批处理重归一化和焦点损失也被纳入以提高准确性。测试结果表明，改进的YOLOv3在工会交叉口（IoU）度量分别为0.5和0.75时，分别具有高达80％和47％的检测精度。它具有ResNet-101的性能优于原始的YOLOv3和两级检测器基于Faster Region-based的卷积神经网络（Faster R-CNN），尤其是对于0.75的IoU度量。测试结果表明，改进的YOLOv3在联合交叉口（IoU）度量分别为0.5和0.75时，分别具有高达80％和47％的检测精度。它具有ResNet-101的性能优于原始的YOLOv3和两级检测器基于Faster Region-based的卷积神经网络（Faster R-CNN），尤其是对于0.75的IoU度量。测试结果表明，改进的YOLOv3在联合交叉口（IoU）度量分别为0.5和0.75时，分别具有高达80％和47％的检测精度。它具有ResNet-101的性能优于原始的YOLOv3和两级检测器基于Faster Region-based的卷积神经网络（Faster R-CNN），尤其是对于0.75的IoU度量。