E2CAV, Pavement layer thickness estimation system based on image texture operators

Resumen (en_US) Context: Public roads are an essential part of economic progress in any country; they are fundamental for increasing the efficiency on transportation of goods and are a remarkable source of employment. For its part, Colombia has few statistics on the condition of its roads; according with INVIAS the state of the roads in Colombia can be classified as “Very Good” (21.1%), “Good” (34.7%), and “Regular” or “Bad” (43.46%). Thus, from the point of view of pavement rehabilitation, it is worth securing the quality of those roads classified as “Regular” or “Bad”. Objective: In this paper we propose a system to estimate the thickness of the pavement layer using image segmentation methods. The pavement thickness is currently estimated using radars of terrestrial penetration, extraction of cores or making pips; and it is part of structural parameters in the systems of evaluation of pavement. Method: The proposed system is composed of a vertical movement control unit, which introduces a video scope into a small hole in the pavement, then the images are obtained and unified in a laptop. Finally, this mosaic is processed through texture operators to estimate the thickness of the pavement. Users can select between the Otsu method and Gabor filters to process the image data. Results: The results include laboratory and field tests; these tests show errors of 5.03% and 11.3%, respectively, in the thickness of the pavement. Conclusion: The proposed system is an attractive option for local estimation of pavement thickness, with minimal structural damage and less impact on mobility and number of operators. Resumen (es_ES) Contexto: Las carreteras públicas son esenciales para el progreso económico de cualquier país, ya que son fundamentales para el incremento en la eficiencia del transporte de bienes, y son una excelente fuente de empleo. Sin embargo, Colombia tiene pocas estadísticas sobre la condición de sus carreteras. e acuerdo con Invías, el estado de las vías pavimentadas colombianas puede resumirse así: 21,1 % son clasificadas como “Muy buenas”; 34,7 %, como “Buenas”, y 43,46 %, como “Regulares” o “Malas”. Entonces, desde el punto de vista de rehabilitación de pavimentos, vale la pena asegurar la calidad de aquellas carreteras clasificadas como “Regulares” o “Malas”. Objetivo: En este trabajo se propone un sistema para estimar el espesor de la capa de pavimento usando métodos de segmentación de texturas en imágenes. Actualmente, el espesor del pavimento es estimado usando radares de penetración terrestre, extracción de núcleos o realizando apiques;