Pavement and base layers local thickness estimation using computer vision

Nowadays, the Colombian paved roads state can be summarized as follows [1]: 21.1% are classified as “Very Good”, 34.7% are classified as “Good” and 43.46% are classified as “Regular” or “Bad”. Then, it is worth noting the need of assuring the pavement rehabilitation quality of public roads, especially those classified as “Regular” or “Bad”. Pavement integral evaluation includes tests such as computing the International Friction Index (IFI), the International Roughness Index (IRI), performing geotechnical and deflectometry studies, and visual inspection of surface defects. In this work, pavement and base layers local thickness estimation based on computer vision methods is proposed. The pavement and base layers thickness estimation are part of geotechnical studies, providing structural properties of flexible pavements. Pavement and base layers thickness were computed using texture filters applied to stitched images acquired from an industrial video-scope. The results reported in this work are focused on comparing between different methods to estimate pavement and base layers thickness.