Low-cost automated image based rutting identification and measurement

Machine-based road surface distress measurements typically involve sophisticated laser and camera systems operated by trained personnel. The high cost of traditional survey methods, however, often results in infrequent surveys of low-volume and urban roads. To address this issue, a low-cost image-based system has been developed in this study. This system is capable of accurately identifying and measuring various distress types such as cracking, potholes, and even water ponding. Rutting measurements, in particular, pose an additional complexity as they require transverse road scanning using multiple lasers. To overcome this, the developed system employs Artificial Intelligence (AI) and computer vision techniques to generate a precise 3D scan from 2D video images, enabling accurate rutting measurements. This paper provides a detailed description of the system, presents the initial validation of the model, and evaluates its accuracy and reliability against actual measurements. The investigation demonstrates promising results in rutting measurement from VIDEO survey.