Abstract:The pavement is vulnerable to damage from natural disasters, accidents and other human factors, resulting in the formation of cracks. Periodic pavement monitoring can facilitate prompt detection and repair the pavement diseases, thereby minimizing casualties and property losses. Due to the presence of numerous interferences, recognizing highway pavement cracks in complex environments poses a significant challenge. Nevertheless, several computer vision approaches have demonstrated notable success in tackling this issue. We have employed a novel approach for crack recognition utilizing the ResNet34 model with a convolutional block attention module (CBAM), which not only saves parameters and computing power but also ensures seamless integration of the module as a plug-in. Initially, ResNet18, ResNet34, and ResNet50 models were trained by employing transfer learning techniques, with the ResNet34 network being selected as a fundamental model. Subsequently, CBAM was integrated into ResBlock and further training was conducted. Finally, we calculated the precision, average recall on the test set, and the recall of each class. The results demonstrate that by integrating CBAM into the ResNet34 network, the model exhibited improved test accuracy and average recall compared to its previous state. Moreover, our proposed model outperformed all other models in terms of performance. The recall rates for transverse crack, longitudinal crack, map crack, repairing, and pavement marking were 88.8%, 86.8%, 88.5%, 98.3%, and 99.9%, respectively. Our model achieves the highest precision of 92.9% and the highest average recall of 92.5%. However, the effectiveness in detecting mesh cracks was found to be unsatisfactory, despite their significant prevalence. In summary, the proposed model exhibits great potential for crack identification and serves as a crucial foundation for highway maintenance.

Sample generation of semi‐automatic pavement crack labelling and robustness in detection of pavement diseases

Asphalt Pavement Crack Identification Based on Two-Stage Training and Multi-Branch Model Integrating Multiple Attention

A Deep Learning Method for Pavement Crack Identification Based on Limited Field Images

DMA-Net: DeepLab With Multi-Scale Attention for Pavement Crack Segmentation

Two-step deep learning approach for pavement crack damage detection and segmentation

Automatic Pixel-Level Pavement Crack Recognition Using a Deep Feature Aggregation Segmentation Network with a scSE Attention Mechanism Module

Automatic identification of pavement cracks in public roads using an optimized deep convolutional neural network model

Virtual generation of pavement crack images based on improved deep convolutional generative adversarial network

Advanced crack detection and quantification strategy based on CLAHE enhanced DeepLabv3+

Automated high-resolution asphalt pavement crack segmentation using deep convolutional neural networks with repeated hierarchical feature fusion

Automated pavement crack detection and segmentation based on two‐step convolutional neural network

A hybrid deep learning pavement crack semantic segmentation

Multi-scale context feature and cross-attention network-enabled system and software-based for pavement crack detection

Weakly supervised pavement crack semantic segmentation based on multi-scale object localization and incremental annotation refinement

Semi-supervised semantic segmentation using cross-consistency training for pavement crack detection

A three-stage pavement image crack detection framework with positive sample augmentation

Automated pavement distress detection using region based convolutional neural networks

Pavement Crack Detection and Segmentation Method Based on Improved Deep Learning Fusion Model

Investigation of pavement crack detection based on deep learning method using weakly supervised instance segmentation framework

Automated highway pavement crack recognition under complex environment

Multi-region Segmentation Pavement Crack Detection Method Based on Deep Learning