Digital image colour analysis to evaluate the moisture sensitivity of hot-mix asphalt mixes

Abstract

Moisture in hot-mix asphalt (HMA) can cause serious damage such as cracks and ruts if left untreated. This is due to the separation of binder and aggregates caused by moisture-induced stripping. Existing approaches for assessing moisture damage in asphalt are time consuming, destructive and expensive to implement. In contrast, digital image analysis provides a simple, inexpensive and non-destructive method for identifying moisture-sensitive mixes. Colour, as an essential characteristic of an image, effectively conveys information about its content. This study seeks to explore the potential of analysing digital image colour in the L*a*b* colour space to assess the moisture sensitivity of HMA after moisture-induced stress tester conditioning. Laboratory-compacted specimens of loose asphalt mixtures from the field were studied by image processing followed by statistical tests of significance. It is found that two parameters – namely, the ‘average colour value for the L* channel’ and the ‘standard deviation value for the b* channel’ – can be used to distinguish between moisture-sensitive and non-sensitive mixes. The findings of this study suggest that future research may find colour to be a helpful criterion for characterising asphalt materials using digital image processing.