Aerosol retrieval from remote sensing image using artificial neural network

The usual method of aerosol retrieval using remote sensing is interpolation of look-up-table (LUT), but it is too time-consuming. However, artificial neural network for nonlinear problem has been not applied widely for aerosol retrieval before. In this paper, aerosol optical depth (AOD) is retrieved using two methods: interpolation and neural network. Then, the retrieval capabilities of the two methods were compared. By comparison, not only is the retrieval error of the neural network within acceptable range, but also it can reduce much processing time. © 2010 IEEE