Using Coastal Aerosol and Cirrus to Dehaze for Landsat 8 OLI imagery

Abstract

Optical image data are often affected by the phenomena of haze and cloud which significantly reduce the quality of images, particularly in remote sensing imagery to identify objects on the ground. Therefore, atmospheric correction to reduce the effects of haze, cloud and shadow is a necessary requirement. Vietnam is located in the tropical monsoon region where is frequently affected by clouds during the rainy season and even in the dry season, thereby the quality of acquired images is reduced. This paper proposes a algorithm to remove the effects of haze for multispectral images of Landsat 8 OLI. This algorithm uses Coastal Aerosol and Cirrus bands to remove the effects of haze on images.