Automated Vision-Based Crack Detection on Concrete Surfaces Using Deep Learning
Automated Vision-Based Crack Detection on Concrete Surfaces Using Deep Learning
复制标题
DOI：
10.3390/app11115229
复制
发表时间：
2021-06-01
影响因子：
2.7
通讯作者：
Kang, Su-Tae
中科院分区：
文献类型：
--
作者：
Rajadurai, Rajagopalan-Sam;Kang, Su-Tae
混凝土结构的裂缝会影响性能，并且是一个主要的耐久性问题。必须及时发现裂缝并进行修复，以保持结构的可靠性和性能。本研究重点研究基于视觉的裂纹检测算法，该算法基于深度卷积神经网络，通过使用迁移学习以更高的分类率对裂纹进行检测和分类。图像数据集由两个后续图像类（无裂纹和裂纹）组成，由 AlexNet 模型进行训练。迁移学习被应用于 AlexNet，包括微调架构的权重、替换两个输出类别（无裂纹和裂纹）的分类层，以及使用随机旋转角度增强图像数据集。经过微调的 AlexNet 模型是通过动量优化器的随机梯度下降进行训练的。精确度、召回率、准确度和 F-1 指标用于评估经过训练的 AlexNet 模型的性能。在学习率为0.0001和6个epoch时，通过训练过程获得的准确率和损失分别为99.9%和0.1%。经过训练的 AlexNet 模型准确预测了 1998/2000 和 3998/4000 验证图像和测试图像，预测精度达到 99.9%。训练后的模型的精确度、召回率、准确度和 F-1 分数分别为 0.99。
Cracking in concrete structures affects performance and is a major durability problem. Cracks must be detected and repaired in time in order to maintain the reliability and performance of the structure. This study focuses on vision-based crack detection algorithms, based on deep convolutional neural networks that detect and classify cracks with higher classification rates by using transfer learning. The image dataset, consisting of two subsequent image classes (no-cracks and cracks), was trained by the AlexNet model. Transfer learning was applied to the AlexNet, including fine-tuning the weights of the architecture, replacing the classification layer for two output classes (no-cracks and cracks), and augmenting image datasets with random rotation angles. The fine-tuned AlexNet model was trained by stochastic gradient descent with momentum optimizer. The precision, recall, accuracy, and F-1 metrics were used to evaluate the performance of the trained AlexNet model. The accuracy and loss obtained through the training process were 99.9% and 0.1% at the learning rate of 0.0001 and 6 epochs. The trained AlexNet model accurately predicted 1998/2000 and 3998/4000 validation and test images, which demonstrated the prediction accuracy of 99.9%. The trained model also achieved precision, recall, accuracy, and F-1 scores of 0.99, respectively.