Abstract
Landscape analysis is usually carried out using a clear sky model, though the importance of landscape under the other weather conditions such as rainy, foggy and snowy conditions is started in many articles and reports. Using computer graphics techniques, it is possible to simulate landscape under such conditions. The fidelity, however, of such simulation images of landscape is not validated on the quantitative basis, which causes difficulties in generating high fidelity landscape images from given weather parameters such as rainfall intensity and relative humidity.
The authors develop a method of generating far-range-landscape images from the weather parameters through organizing and modifying theoretical/empirical formulae on atmosphere/light interaction processes etc.“Weather effects”on far-range-landscape images can be estimated and their fidelity is validated on the quantitative basis. The proposed method is especially useful for practical applications, because many of the weather parameters used here are available from usual meteorological observation data.
