基于贝叶斯分类的两类天气图像自动识别

Automatic Identification of Two Kinds of Weather Image Based on Bayesian Classification

为实现天气现象的自动观测,本文提出了一种基于贝叶斯分类的室外图像的天气现象识别方法,该方法通过提取图像的色相、饱和度和亮度等特征,采用贝叶斯分类,对雾霾和沙尘两类天气现象进行判断。实验结果的准确率达到预期目标,并且能快速、准确地测量室外图像的天气现象。

To realize the automatic observation of weather phenomena, a method for recognizing weather phenomena of outdoor image based on Bayesian classification is proposed. This method extracts image characteristics of hue, saturation and brightness, and adopts Bayesian classification to estimate the haze and sand-dust weather. The accuracy of experiment results reaches the expected aim, and the experimental results show that this method can quickly and accurately measure the weather phenomena of outdoor image.