Crack inspections of automotive engine components are usually conducted manually; this is often tedious, with a high degree of subjectivity and cost. Therefore, establishing a robust and efficient method will improve the accuracy and minimize the subjectivity of the inspection. This paper presents a robust approach towards crack classification, using transfer learning and fine-tuning to train a pre-trained ConvNet model. Two deep convolutional neural network (DCNN) approaches to training a crack classifier—namely, via (1) a Light ConvNet architecture from scratch, and (2) fined-tuned and transfer learning top layers of the ConvNet architectures of AlexNet, InceptionV3, and MobileNet—are investigated. Data augmentation was utilized to minimize over-fitting caused by an imbalanced and inadequate training sample. Data augmentation improved the accuracy index by 4%, 5%, 7%, and 4%, respectively, for the proposed four approaches. The transfer learning and fine-tuning approach achieved better recall and precision scores. The transfer learning approach using the fine-tuned features of MobileNet attained better classification accuracy and is thus proposed for the training of crack classifiers. Moreover, we employed an up-to-date YOLOv5s object detector with transfer learning to detect the crack region. We obtained a mean average precision (mAP) of 91.20% on the validation set, indicating that the model effectively distinguished diverse engine part cracks.
Thanks to the development of deep learning, the use of data-driven methods to detect pavement distresses has become an active research field. This research makes four contributions to address the problem of efficiently detecting cracks and sealed cracks in asphalt pavements. First, a dataset of pavement cracks and sealed cracks is created, which consists of 10,400 images obtained by a vehicle equipped with a highway condition monitor, with 202,840 labeled distress instances included in these pavement images. Second, we develop a dense and redundant crack annotation method based on the characteristics of the crack images. Compared with traditional annotation, the method we propose generates more object instances, and the localization is more accurate. Next, to achieve efficient crack detection, a semi-automatic crack annotation method is proposed, which reduces the working time by 80% compared with fully manual annotation. Finally, comparative experiments are conducted on our dataset using 13 currently prevailing object detection algorithms. The results show that dense and redundant annotation is effective; moreover, cracks and sealed cracks can be efficiently and accurately detected using the YOLOv5 series model and YOLOv5s is the most balanced model with an F1-score of 86.79% and an inference time of 14.8ms. The pavement crack and sealed crack dataset created in this study is publicly available.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.