Pot hole is one of the most common pavement distresses and often appears on poorly constructed pavements. This study proposed a method to rapidly detect and 3D reconstruct the pot hole on the pavement surface using the depth camera. firstly, this study obtained depth images of pot holes with different sizes and degrees by using a depth camera. Secondly, the depth images were imported into the MATLAB to correct impact parameters such as direction, inclination. Third, a 3D reconstruction software was used to reconstruct the pot holes based on the corrected depth images. Forth, the pot holes were filled by plasticine, a plastic material, to obtain the actual 3D model of the pot holes. In the end, the 3D reconstructed result was compared with the actual 3D model. Results of this study can detect the number and size of pot holes in advance, and improve the efficiency of pavement maintenance.

CICTP