Performance analysis and verification of cloud detection algorithm in the presence of various ecological types

Aim of cloud detection is to raise the quality parameters related to surface and atmospheric conditions used in the research area, hence there should be firmer in the cloud detection algorithm. This article includes a cloud detection algorithm based on thresholds which is used to differentiate clouds from underlying surface applied to the visible band. Four ecological types: Soil, Coastal, Land and Water have been referred in an algorithm to ensure its validity. By identifying strengths and weaknesses of different algorithms, analysis and verification of proposed approach carried on NOAA and VIRR data sets(each consist of approximately 300 images). Analysis and verification of proposed approach carried on NOAA and VIRR data sets and a comparison with an official cloud masking of VIRR and an existing algorithm: A daytime cloud detection algorithm for FY-3A/VIRR data are included; the test result describes that this algorithm gives better performance than an existing approach.