关键词: 沥青路面, 健康光谱指数, Worldview-2, 光谱特征, 沥青老化

Abstract:

The rapid development of highway transport networks has increased much work load to road maintenance departments in China, and consequently it is currently a pressing demand to develop new technical support to rapidly and accurately collect the health conditions of road pavements. In contrast to the conventional methods, previous research indicated that remote sensing might offer a new approach for the monitoring of pavement conditions of highway roads. This paper first tends to examine the spectral responses and features of the road pavements with different aging conditions and pavement materials based on field measurements of the pavement spectral reflectance. In addition, this study also tries to construct effective road pavement condition index from satellite remote sensing data for the monitoring of pavement health conditions. One of the findings shows that the slope of the spectral curves in the wavelength region of 400~900 nm grows bigger from negative to positive with the gradual aging of the asphalt pavements based on the field measurements of the pavement spectra. After that, several spectral index models were built up to monitor and evaluate the road pavement aging conditions by means of simple arithmetical calculation such as ratio and normalization. To demonstrate the applicability of the proposed indices to satellite remotely sensed data, a Worldview-2 image acquired on September 21, 2013 in the Liangxiang area near the sixth Ring Road south, Beijing City was used to analyze the road pavement health conditions, and to verify these models using the Munsell Scale Card values that were collected together with the spectral measurements and used to indicate the pavement aging conditions in our study. The image was first preprocessed in RSI ENVI software package, such as radiometric and geometric corrections, and subset. The four proposed indices were calculated from the Worldview-2 image in the study area and evaluated using the in-situ measurements of the pavement health conditions by visually comparing the performance of the proposed spectral indices in characterizing the asphalt pavement aging conditions. Furthermore, correlation analysis between spectral pavement condition indices and the Munsell Scale Card values shows that the logarithmic health index can achieve the biggest determinant coefficient （R²=0.72,n=23） and may be applicable in the monitoring of road pavement conditions. The case study indicates that road pavements with different materials and aging conditions have distinct spectral response in visible and near-infrared wavelength, and satellite remote sensing can be employed in the rapid mapping and assessment of large-range asphalt road pavement conditions. Thus, the study extends remote sensing applications and offers a new technique for the road maintenance departments. Future work may explore spectral mixture analysis in asphalt condition mapping from low-altitude unmanned aviation vehicle （UAV） hyperspectral imagery.

Key words: asphalt pavement, spectral index, high-resolution remote sensing, spectral signatures, asphalt aging