Dynamic Pothole Detection: A Six-Month Study of Depth Variation and Characterization Using Scatter Plot Techniques

Abstract Potholes represent a significant threat to road infrastructure, contributing to road safety risks and escalating maintenance costs. This research investigates dynamic pothole detection and depth variation over a six-month period, employing a combination of RGB value analysis, pothole depth measurements, scatter plot visualization with the help of terrestrial laser Scanner, and artificial intelligence (AI) techniques. The study aims to develop an automated model that characterizes potholes not only in terms of their spatial location but also by assessing their depth and temporal progression. RGB values from high-resolution images of road surfaces are analyzed to detect pothole features, while depth measurements are recorded using advanced sensors. AI algorithms, including image processing and machine learning, are applied to process the data, detect potholes, and analyze their development over time. Scatter plot techniques are utilized to visualize the correlation between pothole depth, location, and environmental factors, providing deeper insights into the conditions that drive pothole formation and expansion. The results highlight the potential of integrating image processing and AI in proactive road maintenance strategies, offering an innovative approach to early pothole detection and characterization. This model could significantly enhance predictive maintenance, reduce road repair costs, and improve overall road safety.