Investigation on Aggregate Adhesion Segmentation Method in an Asphalt Concrete Digital Image

Abstract

The processing method of asphalt mixture digital images was investigated on the basis of the gray – level distribution of asphalt concrete images. Some bond areas were found between different aggregates. The watershed segmentation method was used to separate aggregates, the results of which were verified. Aggregate orientations in binary images were analyzed statistically. The correlation between aggregate orientation and aspect ratio was determined. Results showed that the watershed segmentation method can separate aggregates, and that aggregate orientations are affected by the aspect ratio. A more oblate aggregate corresponds to a more concentrated placed orientation.