Abstract Modern asphalt microstructural analysis often requires samples with minimal thickness and high smoothness. The unavoidable coffee-ring effect during evaporation leads to a nonuniform deposition pattern, resulting in thickness differences between the edge and center regions of the sample. Characterizing the microstructure and morphology of asphalt is essential but challenging because of its limited optical properties. This study presents an innovative image processing method to analyze the thickness distribution of asphalt film samples prepared through drop evaporation. A precise quantitative relationship between film thickness and grayscale values is established using angle-varying tests, enabling the generation of thickness distribution maps based on grayscale data. Image segmentation algorithms divide the sample into edge, transition, and center regions, allowing for detailed regional analysis. The mean absolute deviation (MAD) algorithm is used to quantify smoothness in each region. Results show that although the center region is thinner, it has lower smoothness than the transition zone. This suggests that the transition zone is optimal for sampling because it meets the requirements of minimal thickness and high smoothness. In summary, this image analysis technique identifies optimal regions on asphalt samples for obtaining high-quality specimens, thereby supporting sampling requirements in other asphalt studies. This study is notable for its (1) novel grayscale analysis method for determining asphalt film thickness and smoothness, (2) image segmentation and MAD algorithms that identify optimal sampling regions in asphalt films, and (3) method that addresses nonuniform deposition from the coffee-ring effect in asphalt analysis.
Wei et al. (Thu,) studied this question.