多数据源融合的高速公路气象识别方法

Highway Meteorological Recognition Research of Multiple Data Fusion

对与高速公路行车安全有重要影响的雨、雪、雾、路面结冰、路面霜冻、路面高温等环境气象和路面状况进行实时监测与预报。通过现场多传感器及摄像机获取监测点气象环境原始数据,对摄像机图像进行灰度变换、二值化处理、形态学运算,并将图像处理结果与多传感器数据融合,进行加权平均运算,分析得出气象识别结果;在处理中心,将气象识别结果和当地气象特征相结合,对每一种气象识别结果再进行成因推理、分析,得出现场检测点处的环境气象和路面状况。给出了雨、雪、雾天气识别的图像处理方法和气象特征综合识别方法,实现以较少的传感器数量获取更准确、更详细、更完整的道路气象特征的目标。

Real-time monitoring and forecasting the environmental weather such as rain,snow,fog,icy,etc and road conditions such as road frost,road temperature and other meteorological elements for the highway have a major impact on highway traffic safety. We got the original data for meteorological environment of monitoring points through the multi-sensor and camera at the scene,and then we did the camera image gray-scale transformation,binary processing,and morphological operations. And we merged the result of image processing with multi-sensor data to have the weighted average calculation,and then we got the the meteorological recognition result. In the dealing centers,we combined the recognition results of weather with local meteorological characteristics and then analyzed the recognition results for each meteorological causes and did further reasoning,at last we got the weather androad surface conditions of testing point. The comprehensive identification method of meteorological characteristics and the image processing method of the identification of rain,snow,and fog were put forward,which achieves the result of obtaining more accurate,more detailed,more complete road meteorological characteristics with less number of sensors.