Development Of Landsat Cloud Free Image Data For Classification Of Land Cover-Case Study In Dak Lak Province

Abstract

Land cover mapping by optical remote sensing has many obstacles including clouds. Clouds block solar radiation coming to earth surface and reflective radiance from the earth surface to remote optical sensors resulting. Therefore, clouds result no-signal areas in images that cannot be used for study of ground objects. In many cases, thin clouds degrade quality of reflective radiance and some times alter, unexpectedly, spectral reflectance characteristics of ground objects leading to false classification. In this paper, the authors present an algorithm on application of multidate for development of cloud free image. The used image data were received in rainy and dry seasons and by stacking, cloud free images representing rainy and dry seasons were created. These cloud free images can be used further for classification of land cover in rainy and dry seasons. Experiments were conducted with Landsat 8 OLI images with path/row number 124/51 covering Dak Lak province of Vietnam. The results of case study were development of cloud free image data representing rainy and dry seasons allowing separation of evegreen and deciduous forests in the study site.