Study on Damage to Flexible Pavement Road in Thailand: Identification Types of Crack on Roads Using Image Processing Takes

The detection and identification of flexible pavement road cracks pose a significant challenge for civil engineers when they start work at the beginning level, as it requires extensive work experience of engineers in classifying types of cracks for planning road maintenance. Inaccurate analysis and identification of cracks can negatively affect the quality of maintenance work and threaten the safety of road users in the future. This paper presents a case study of 60 asphalt roads managed by the Provincial Administrative Organization (PAO) in central of Thailand (case of two provinces, Uthai Thani and Nakhon Sawan), focusing on identification type road cracks. The study aims to address issues related to data preparation and classification across various domains Our data are fed model, which is photography to medium-resolution images of flexible pavement road cracks so that guidelines for engineers' inspection and identification of road cracks. We propose the development of a U-Net model specifically designed for image processing tasks. The results indicate that the U-Net model is effective in converting road cracks images into segmentation masks, and this effectiveness is evaluated using four well-defined key performance indicators.