Key points are not available for this paper at this time.
The efficiency of transportation and road safety are critical factors for economic well-being, and they are substantially influenced by the condition of road surfaces. Unfortunately, existing practices often result in delays of weeks or even months before government authorities address road surface damage and defects. This delay primarily stems from a lack of timely awareness regarding such issues. Pavement distress, such as cracks, potholes, and surface deterioration, poses significant risks to road users and can lead to costly infrastructure damage. Traditional methods of pavement assessment are often labor-intensive and time-consuming, making them impractical for large-scale road networks. Deep learning, within the realm of artificial intelligence, has risen as a potent and influential tool for automating the detection and recognition of pavement distress. This paper examines the most advanced deep learning models in the current state of the field, datasets, and evaluation methods used in pavement distress analysis.
Sheeja et al. (Thu,) studied this question.