A review of the progress in machine vision-based crack detection and identification technology for asphalt pavements

With the aging of transportation infrastructure and the increasing frequency of use, the detection and identification of cracks in asphalt pavements are crucial for ensuring road safety and maintenance efficiency. Traditional manual inspection methods are not only inefficient and limited in accuracy but also susceptible to subjective factors and environmental conditions. In contrast, machine vision-based crack detection technology enhances the efficiency and reliability of detection through automated image acquisition and analysis processes. This article reviews the latest advancements in machine vision-based crack detection technology for asphalt pavements, with a particular focus on the applications of digital image processing and deep learning. Although image processing-based methods perform well in detecting cracks against simple backgrounds, they exhibit poor robustness under complex lighting and background conditions. On the other hand, deep learning-based methods, while effectively handling complex image data, rely on large amounts of annotated data and significant computational resources. Through critical analysis, the article evaluates the strengths and weaknesses of existing technologies and looks forward to future research directions that integrate multiple sensing data and automated data annotation tools, aiming to further advance and innovate road maintenance technology.