沥青混合料图像多阈值分割方法

将基于遗传策略的多阈值分割法应用到沥青混合料图像分割中,解决了边缘检测法、单阈值法抗干扰性差、分割结果不能直接用于级配分析的难题。实验结果表明,该方法的分割精度优于边缘检测类和单阈值法,而且可将沥青混合料中的粗骨料、细骨料、纹理和空隙等组成部分分离出来,为后续的级配分析计算奠定基础,具有广阔的应用前景。

Image of HMA Multi-level Threshold Segmentation Method

A image of hot mix asphalt (HMA) multi-level threshold segmentation method based on genetic strategy is proposed, compared with Roberts operator, Canny operator and Ostu method in the present literature. Experimental results show that the method has higher segmentation precision than others, and HMA is separated into coarse aggregate of concrete, fine aggregate concrete, interstice and traces, which provides a basis for gradation analysis, and has great application prospects.