Robust urban road surface monitoring system using Bayesian classification with outlier rejection algorithm

An urban road surface monitoring system is developed to monitor highway conditions. This paper uses principle component analysis (PCA) that combines 94 GHz dual-channel polarimetric radiometer and automatic weather station measurements. The goal of this paper is to combine these observations into a smaller number of variables through PCA. Each new variable has a unique physical interpretation. In particular, we are concerned with principal component variables whose weightings include brightness temperature and road surface temperature. Then, four different road surface classes (dry, wet, snowy, and icy) are classified by Bayesian classification. The novelty of our approach relies on using 94 GHz dual-channel polarimetric radiometer is used to investigate behavior of the brightness temperature of different road surface conditions in an open-air laboratory. On the other hand, the median absolute deviation algorithm is used to remove extreme values automatically from a given database of weather information (including brightness temperature, road surface temperature, wind speed, etc.) for robust classification. Finally, the road surface statuses were found to be well classified by the proposed method in real time.