Pavement structure segmentation method based on results derived from ground-penetrating radar data

This paper presents a pavement structure segmentation method based on the results derived from ground-penetrating radar (GPR) data. Surface type, layer thickness, base/sub-base type, shortest segment and outlier point criteria are developed to perform the segmentation. A boundary comparing method using t-tests is developed to perform layer thickness criteria. The value t=4 is suggested for the HMA (hot mix asphalt) or aggregate layer; and t=6 is suggested for Portland cement concrete and cement-treated base. During segmentation, segments that are shorter than the minimum length of 100m are merged into adjacent segments. Furthermore, the flow charts of pavement structure segmentation as well as a procedure for automated segmentation are detailed. Proposed rules for fixing segment boundaries between GPR observations are also presented. Several segmentation examples are calculated by the automated procedure using the data from a pilot project in California. The results show that the method presented in this paper is reasonable for pavement structure segmentation based on GPR data.