3D pavement macrotexture parameters from close range photogrammetry
Marcelo S. Medeiros, Lucas Feitosa de Albuquerque Lima Babadopulos, Renan Santos Maia, Verônica Teixeira Franco Castelo Branco
Abstract
The 3D modelling of pavement surface using digital close-range photogrammetry (CRP) was thoroughly analyzed in this work. The accuracy and precision of the technique were effectively measured. A digital 3D model generated by CRP was compared to the model generated by a laser scanner, considered here as a benchmark for precision. The precision, measured as per cent deviation from a computer-generated model, considered here as the true value, was found to be 3.4% and the accuracy, measured as per cent error, was found to be 4.2%. A study involving 85 locations and 5 different surface types was also conducted to compare the calculated macrotexture parameters to those obtained from the Sand Patch Test and British Pendulum Test (BPT). The creation of a 3D surface model allowed the investigation of parameters other than the averages of height and depth proposed in the current standards. Moreover, the process described in this work can be automated and with proper standardization could be used for management and decision making at the infrastructure's network level. The British pendulum did not show a high correlation to the 3D macrotexture parameters except for the Mean Texture Depth (MTD) calculated from the sand patch itself (R2 = 59.7%).