Reconstruction of 3D Pavement Texture on Handling Dropouts and Spikes Using Multiple Data Processing Methods

Tire–pavement interactions, like friction and rolling resistance, are significantly influenced by pavement macro-texture and micro-texture. Accurate texture measurement at the micro-texture level is vital to achieve the desired level of safety, comfort, and sustainability of the pavement. However, the existence of dropouts and spikes in the collected data is still inevitable based on current laser devices, which leads to erroneous texture characterization. This study utilized an advanced laser sensor to measure three-dimensional (3D) pavement texture at the micro-level at a given speed. Using a proposed interpolation method, the dropout areas in the raw measurements were filled up. Butterworth’s high-pass and low-pass filters were applied to separate two texture components from the profile. Based on a statistical analysis for the micro-texture amplitude, an appropriate threshold was determined in order to identify the spikes. A three-step-spike-removal method was proposed and found to be effective in clearing the spikes. The 3D pavement profiles were finally reconstructed without dropouts and spikes. Mean profile depth (MPD) was calculated with different baselines. It was found that the presence of spikes leads to a greater MPD value and the MPD is sensitive to the baseline length. A shorter baseline is recommended to mitigate the impact of spikes on the accuracy of the MPD.