Spatiotemporal matching method for tracking pavement distress using high‐frequency detection data

Abstract

Various algorithms based on deep learning have achieved promising results in pavement distress detection. However, the detected distresses are not tracked throughout the life cycle. In long‐term application scenarios, pavement distresses may take on different forms due to image acquisition mode, distress development, and environmental change, which make tracking distresses a tough question. We present in this study a spatiotemporal matching method based on high‐frequency real pavement distress datasets. Pavement distresses of fixed routes were collected 30 times over 5 months, and distresses with spatiotemporal information were obtained at time series. We apply image rectification, stitching and distress class, and bounding box generation algorithms for pre‐processing to align the collected images to the same‐detail level and angle. A four‐step spatiotemporal matching module is designed, including global positioning system (GPS) filtering, class filtering, relative position filtering, and distress feature filtering. The results reveal that the comprehensive rank‐3 hit rate of the matching method reaches 88.73%, and the method is robust to environmental factors, which helps show performance decay of distresses and the effect of maintenance operations. It is concluded that the spatiotemporal matching method is convenient to operate, and it lays the foundation for an agency to track distress evolution and make timely treatment of distresses in the life cycle.

A controllable generative model for generating pavement crack images in complex scenes

“…In recent years, advanced machine learning algorithms have been applied in the field of civil engineering, especially deep learning algorithms (Alam et al., 2020; Rafiei & Adeli, 2017), which have played a significant role in the prediction of material properties (Pereira et al., 2020; Rafiei et al., 2017), assessment of engineering benefits (Martins et al., 2020; Rafiei & Adeli, 2018), and detection of pavement defects. Automatic detection methods and equipment for pavement crack images based on deep learning have been developed (Pan et al., 2023). However, these commonly used deep learning methods often face difficulties in detecting cracks in complex scenes, which include images with complex cracks, brake marks, water marks, and shadows (Bai et al., 2021; Yan & Zhang, 2021).…”

AI‐enabled airport runway pavement distress detection using dashcam imagery

“…Deep convolutional neural networks (DCNNs) can surpass traditional techniques and have been applied to numerous civil infrastructure health monitoring applications (Amezquita‐Sanchez et al., 2020; X. Pan et al., 2023; Rafiei & Adeli, 2017a, 2018a, 2018b), and more specifically, pavement distress classification, detection, and segmentation (Pauly et al., 2017; Urban et al., 2017; L. Zhang et al., 2016). Recent advances like transfer learning improve results by leveraging generic features from large datasets (S. J. Pan & Yang, 2010).…”

“…Other state‐of‐the‐art approaches utilize DCNNs with techniques like PCM, fully connected neural networks, fully convolutional networks, U‐Net, and residual networks (Bang et al., 2018; J. Chen & He, 2022; Cheng et al., 2018; Dung & Anh, 2019; Fan et al., 2018; X. Wang & Hu, 2017; Yang et al., 2018) and transformers (Tong et al., 2023). Furthermore, recent advancements, such as spatiotemporal matching for tracking evolution over time (N. Pan et al., 2023), cross‐scene transfer learning for enhanced adaptability (Y. Li et al., 2021), pixel‐level multi‐distress detection for precision (A. A. Zhang et al., 2022), and a fusion model for comprehensive crack identification (B.…”