Research on the Classification Method of Complex Snow and Ice Cover on Highway Pavement Based on Image-Meteorology-Temperature Fusion

Precise classification of ice and snow cover on winter pavement is crucial for safe driving and efficient road maintenance, but the classification performance of the existing methods is insufficient in the face of the increase in the detailed classification status of complex ice and snow cover. Therefore, this article proposes an image-meteorology-temperature fusion-based roadway snow and ice cover classification method to better extract the differential features of complex snow and ice cover through the incorporation of new methods and the better combination of different methods on the basis of the existing detection methods. Conduct four classification experiments of snow and ice cover with different classification methods and classification experiments when the complexity of the cover increases, and then evaluate the detection performance of the different methods. The experimental results show that, without considering the temperature method (TM) and meteorological method (MM) with lower classification precision, the method proposed in this article, compared with other methods, can improve the average precision (AP) by 2.1%–11.5% for four classifications of snow and ice cover; the decline rate of AP can be reduced by 2.1%–11.5% for the increase of the three classifications of snow and ice cover to the four classifications; and the decline rate of AP can be reduced by 0.9%–15.6% for the increase of the four classifications of snow and ice cover to the five classifications. The above experimental results prove that the method proposed in this article has a better performance in precision and robustness in the classification of complex snow and ice cover.