基于图像处理的再生沥青混合料体视学分析

Stereological analysis of recycled asphalt mixture based on digital image segmentation method

针对目前再生沥青混合料新旧集料混合不均而造成混合料性质不稳定的现象,文中提出采用化学染色法区别新旧集料、对比分析多种阈值的分割方法。通过图像方法分析再生沥青混合料的不同体积组成,在此基础上提出基于Laplace的数字图像分割处理方法,结合体视学分析,实现三维实体转换二维数字图像,从而进行再生沥青混合料的级配设计。结果表明,化学染色法结合Laplace图像分割处理方法可以还原初始图像中的不清晰细节,增强图像整体对比度;通过体视学分析提出合适的骨料修正系数,有效分析再生混合料的级配组成,为完善再生沥青混合料设计提供参考。

Uneven mixing of old and new aggregates in recycled asphalt mixture may lead to unstable performances of mixture. A dyeing binding method （ DBM ） is proposed in order to distinguish recycled and virgin aggregates in the recycled asphalt mixture. Some threshold segmentation methods are comparatively analyzed to describe different results of volumetric components of the recycled asphalt mixture. Then a digital image segmentation method （ DISM） based on Laplace method is proposed. Through this method and stereological an a ly sis ,the three-dimensional view can be converted into the two-dimensional digital images. The design of recycled asphalt mixture can be accomplished with the help of this method. The results indicate that blurred details of original images can be reduced and the contrast of entire images can be strengthened by using the DBM and DISM. Based on the s tudy, appropriate aggregate correction coefficient is provided by stereology analysis. The DBM/DISM and stereology analysis can provide effective mix design of recycled asphalt mixture.