Study On Digital Image Features Of Asphalt Concrete

Asphalt concrete is a non-homogeneous material that has anisotropic property, nonlinearity and variability. To understand the property or to control the quality need investigate to the microstructure of asphalt. Digital image processing provides a way to acquire and describe the microstructure of asphalt concrete, which can directly acquire shape features and spatial distribution of aggregates.The asphalt image acquired can be analyzed only after restoration and segmentation. Segmentation is one of the most difficulty tasks in image processing. The performance of segmentation determines the reliability of subsequent handling. The multilevel segmentation algorithm which concerned edge detection and histogram has been proved a robust tool fit to the asphalt concrete image.The image from asphalt concrete cutting surface has three aspects of features, which include shape properties, structure characteristics and holes distribution. By multivariant regression we can find the linear relation between shape properties and stability, shape properties and flow value, shape properties and skin drying density, structure characteristics and stability. By curve estimate we can find the logarithm relation between the accounts of holes and VV, cubic relation between the accounts of holes and flow value. We can improve the accuracy of accounting of holes by pattern classifying of holes. The image of road surface has its own characteristics that match the texture category of digital image processing. The road surface's texture features and road's micro depth have obvious linear relation.We can trace the displacement of asphalt concrete surface when deforming through the CCD equipment, and get the strain field distribution by interpolation, that is a powerful tool to analyze the microstructure of asphalt concrete.