Mean profile depth of pavement surface macrotexture using photometric stereo techniques

Pavement surface texture affects many vehicle and road characteristics; therefore, efforts are needed to develop more advanced techniques for evaluating pavement texture. Photometric stereo techniques are widely used to recover three-dimensional shapes of objects and can also be used to recover surface texture. A four-source photometric stereo system for recovering pavement surface texture is proposed. Five types of pavement surfaces were tested to validate the system. Mean profile depths computed from the recovered surface were compared with those measured manually by using a depth dial gauge. The sensitivity of the technique to illumination angle was studied for five zenith angles: sigma=26, 28, 30, 32, and 34 degrees. The computed mean profile depths from the proposed system are linearly correlated with those from the depth dial gauge with coefficients of determination ranging from 0.82 at sigma=34' to 0.92 at sigma=30 degrees. Tests showed that the proposed system can be used to recover pavement surface heights and to estimate the mean profile depth.