Study on the Influence of Asphalt Pavement Texture on Drainage Performance

Abstract

During rainy weather, water films on road surfaces can reduce skid resistance, leading to hydroplaning and traffic accidents. This study involves collecting images of asphalt pavement and processing these images using image processing techniques to obtain the texture occupancy of the recessed areas of the pavement. To eliminate noise in the asphalt pavement images and make edge features more distinct, the preprocessing steps include grayscaling, image de-noising, and image enhancement. Threshold segmentation is then used to obtain binary images of the asphalt pavement, and contour detection algorithms are applied to extract the texture distribution of the pavement. This allows for the calculation of the texture occupancy ratio of different approximated recessed areas. Water film thickness measurements were performed using a laser refraction method under various texture configurations and rainfall intensities. The study developed predictive models for water film thickness based on the texture occupancy ratio of different approximate diameters. Analysis of the predicted values from these models compared to actual measurements revealed that the predictive model for textures with an approximate diameter greater than 15mm is reliable for estimating water film thickness. Additionally, in the image processing phase, the texture occupancy ratio for textures with an approximate diameter greater than 15mm resulted in faster image processing speeds and was less susceptible to noise interference.