Identification and Quality Classification of Pavement Road Markings.
This is an AI research project using computational neural nets to identify the characteristics of road markings and quantifying their quality (wear).
This will allow the identification of the type of marking and classify the shape and wear patterns of center road markings (such as lines, arrows, symbols and text messages) that have been damaged by traffic & weather. These road markings/symbols when heavily damaged or worn away, cannot be classified using conventional machine vision techniques. This will also allow our systems to measure the amount of retroreflective material coverage that is still present/missing on the markings. Classifying degraded road markings was mentioned as a CEDR requirement in the CEDR Premium Project.