Road Surface Classification Based on Millimeter-Wave Radar Backscatter Measurements

This paper investigates the backscattering characteristics of millimeter-wave radar for road surface classification and evaluates the performance of three classification algorithms using a six-class road surface dataset. We propose a novel set of features and implement them on the three machine learning algorithms for classification. Our research innovation lies in achieving road surface classification with fewer features in a more realistic experimental setup. The results demonstrate the effectiveness of the selected features in accurately classifying road surfaces.