Intelligent Detection Method for Concrete Dam Surface Cracks Based on Two-Stage Transfer Learning

Abstract

The timely identification and detection of surface cracks in concrete dams, an important public safety infrastructure, is of great significance in predicting engineering hazards and ensuring dam safety. Due to their low efficiency and accuracy, manual detection methods are gradually being replaced by computer vision techniques, and deep learning semantic segmentation methods have higher accuracy and robustness than traditional image methods. However, the lack of data images and insufficient detection performance remain challenges in concrete dam surface crack detection scenarios. Therefore, this paper proposes an intelligent detection method for concrete dam surface cracks based on two-stage transfer learning. First, relevant domain knowledge is transferred to the target domain using two-stage transfer learning, cross-domain and intradomain learning, allowing the model to be fully trained with a small dataset. Second, the segmentation capability is enhanced by using residual network 50 (ResNet50) as a UNet model feature extraction network to enhance crack feature information extraction. Finally, multilayer parallel residual attention (MPR) is integrated into its jump connection path to improve the focus on critical information for clearer fracture edge segmentation. The results show that the proposed method achieves optimal mIoU and mPA of 88.3% and 92.7%, respectively, among many advanced semantic segmentation models. Compared with the benchmark UNet model, the proposed method improves mIoU and mPA by 4.6% and 3.2%, respectively, reduces FLOPs by 36.7%, improves inference speed by 48.9%, verifies its better segmentation performance on dam face crack images with a low fine crack miss detection rate and clear crack edge segmentation, and achieves an accuracy of over 85.7% in crack area prediction. In summary, the proposed method has higher efficiency and accuracy in concrete dam face crack detection, with greater robustness, and can provide a better alternative or complementary approach to dam safety inspections than the benchmark UNet model.

Vison Transformer-Based Automatic Crack Detection on Dam Surface

“…This approach addresses the issue of inadequate global information in low-dimensional features, thereby enhancing the prediction of crack direction. In addition, the encoding-decoding structure [26,[29][30][31]36] is a commonly employed design in semantic segmentation networks, where encoding corresponds to feature extraction, and decoding involves resolution recovery. Moreover, skip connections [16,26,32] are implemented between the encoding and decoding layers of the identical resolution to mitigate the loss of fine-grained details and boost crack segmentation.…”

“…Within the domain of crack detection research, the majority of crack segmentation tasks are predominantly executed through the utilization of encoder-decoder architectures [26][27][28]. Various wellknown architectures and their improvements such as UNet [29,30] and DeepLabv3+ [31,32] were also proposed to conduct dam crack detections. These deep learning-driven semantic segmentation methodologies have been found to offer enhanced detection results under challenging and noisy environmental scenarios, and precision measurement in cracks [27,31].…”