|
|||||
|
|
| 黄渤海晴空像元下雾霾识别的机器学习算法 | |
| 作者姓名： | 沈晓晶 姜祝辉 陈 建 刘 娟 马 明 |
| 作者单位： | 北京应用气象研究所 |
| 摘 要： | 为实时有效地识别和监测雾霾，选取2013年10月4日与10月5日的两景MODIS遥感图像，利用不同类型像元的光谱差异，基于神经网络模型，建立海上雾霾的遥感识别机器学习算法。经检验表明，在晴空条件下，机器学习算法反演精度相对较高，已超过98%，利用算法可以实现对中国近海上空雾霾的监测。 |
| 关 键 词： | 雾霾 机器学习 晴空 |
Machine learning algorithm of haze detection over the Yellow and Bohai Sea |
|
| Authors: | SHEN Xiaojing JIANG Zhuhui CHEN Jian LIU Juan MA Ming |
| Affiliation: | Institute of Applied Meteorology in Beijing |
| Abstract: | In order to identify and monitor the haze effectively, in this paper, we selected two MODIS images on 4th and 5th October, 2013. Based on spectral differences between different kinds of images and neural network model, a machine learning algorithm for identification of marine haze was established in this paper. By validation, the accuracy of the algorithm was more than 98%. With this haze detection algorithm, we can monitor haze for long time over East China Seas. |
| Keywords: | Haze Machine learning Sunny |
| 点击此处可从《》浏览原始摘要信息 | |
| 点击此处可从《》下载全文 | |
