Rapid Highway Meteorological State Classification Method Based on LBP Features
The discretization of characteristic signals in point cloud polarization may cause large classification error.A rapid classification method of highway meteorological state based on LBP features is proposed.According to the matrix filling meth-od,the missing polarization parameter data are reconstructed to obtain the multi-scale information of the image.The multi-scale LBP is used to obtain the spatial structure of the contained image texture and strengthen the image representation ability.The Doppler weather radar of point distribution is used to filter the echo intensity of the road weather state,and obtain the polar co-ordinate scanning results of the road weather state radar.The Bayesian network classifier is used to calculate the maximum pos-terior probability,divide the value range of meteorological state categories,and complete the rapid classification of meteorologi-cal states.It can be seen from the experimental results that the classification error of this method is less than 0.30,which a-chieves accurate classification.