基于LBP特征的公路气象状态快速分类方法  

Rapid Highway Meteorological State Classification Method Based on LBP Features

作　　者：谢克勇 岳旭 谢佳杏 王嘉琦 王明飞 袁崇斌 XIE Keyong;YUE XU;XIE Jiaxing;WANG Jiaqi;WANG Mingfei;YUAN Chongbin(Meteorological Service Center of Jiangxi Province,Nanchang 330096,China;Climate Center of Jiangxi Province,Nanchang 330096,China)

机构地区：[1]江西省气象服务中心,江西南昌330096 [2]江西省气候中心,江西南昌330096

出　　处：《微型电脑应用》2024年第5期41-44,共4页Microcomputer Applications

基　　金：江西省03专项及5G项目(20212ABC03A029)。

摘　　要：由于点云偏振存在特征信号离散化现象,造成分类误差较大的问题,因此提出基于LBP特征的公路气象状态快速分类方法。根据矩阵填充方法对存在缺失的偏振参量数据进行重构,获取图像多尺度信息。利用多尺度LBP获取蕴含图像纹理的空间结构,强化图像表征能力。使用点状分布的多普勒气象雷达筛选公路气象状态回波强度,获取公路气象状态雷达的极坐标扫描结果。使用贝叶斯网络分类器计算最大后验概率,划分气象状态类别取值范围,完成气象状态快速分类。由实验结果可知,该方法分类误差低于0.30,可以实现精准分类。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 method,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 coordinate scanning results of the road weather state radar.The Bayesian network classifier is used to calculate the maximum posterior probability,divide the value range of meteorological state categories,and complete the rapid classification of meteorological states.It can be seen from the experimental results that the classification error of this method is less than 0.30,which achieves accurate classification.

关 键 词：LBP特征 公路气象 快速分类 矩阵填充法 

分 类 号：TN959.4[电子电信—信号与信息处理]