Optical profilometry for road normal distribution measurement at a large scale: current issues and data interpretation

Accurate photometric modeling of road pavements requires knowledge of the normal distribution function (NDF), which characterizes the statistical orientation of surface microfacets. Direct measurement of the NDF on real pavements is rarely feasible; instead, we propose estimating it from topographic data acquired via laser profilometry. This paper develops the theoretical framework to compute the NDF and assess measurement quality through an inter-line correlation indicator. Experimental measurements on various road samples are performed using both laser profilometry and photometric stereo imaging for validation. Results show that anisotropies and acquisition deviations in profilometric data can strongly affect NDF estimation, while the proposed correlation analysis efficiently reveals such artifacts. This study establishes a practical methodology for evaluating surface normal distributions over extended road areas, paving the way for realistic large-scale photometric simulations of urban environments.