Geometric Evaluation of Road Signs Using Radiometric Information from Laser Scanning Data

Road signs appear as one of the most important elements for safety purposes and their inspection is typically included in the road management systems. Nowadays, it is becoming more common the use of high performance vehicles to improve the efficiency of the road inspection works. They include sensors as laser scanning for geometric measurements, optical profilers for the detection of cracks on pavement, ground penetrating radar for measuring pavement thickness and video cameras for the visual reference. The road signs use reflective paints to maximize the visibility by the drivers. This coating produces a high reflected radiation in the photoelectric detector of the laser scanners, which allows establishing a simple filter for the 3D segmentation of the road sign. In this work, a number of road signs are evaluated using a laser scanner Riegl LMS Z390i. The geometrical parameters under study were the sign flatness and the tilt angle with the plane of the pavement. A Matlab algorithm is developed for all the data processing (3D segmentation and evaluation of geometric parameters). The algorithms developed open the possibility to use laser scanner retroreflectivity for segmentation purposes in the evaluation of the condition state of road signs and could increase the productivity of the inspection works.