Fig. 1. Extraction process of rut 3D contour information

Fig. 2. Model enhancing the grooving characteristic process of ruts. (a) Cross section of point cloud after pretreatment; (b) cross section of the relative depth model of road point cloud

Fig. 3. Process of building pavement point cloud relative depth model. (a) Primary point cloud; (b) point cloud after pretreatment; (c) pavement point cloud relative depth model

Fig. 4. Rut boundary point and bottom point

Fig. 5. Boundary point of rut curvature

Fig. 6. Immersion process of point cloud watershed algorithm

Fig. 7. Rut depth calculation

Fig. 8. Results of rut contour extraction. (a) Heavy rut; (b) light rut

Fig. 9. Boundary accuracy. (a) Precision of left edge position of left rut; (b) precision of right edge position of left rut; (c) precision of left edge position of right rut; (d) precision of right edge position of right rut

Fig. 10. Measured rut depth using different methods. (a) Left rut depth of heavy rut; (b) right rut depth of heavy rut; (c) left rut depth of light rut; (d) right rut depth of light rut