Abstract
In this paper, an approach for automatic cloud detection and localization in satellite remote sensing images is introduced. Cloud detection is useful in improving the accuracy of land cover classification in cloudy satellite images. The accurate detection of clouds in satellite images is vital for many atmospheric and terrestrial applications. In this paper we propose an algorithm for automatic cloud detection based on neutrosophic set and wavelet transform. The proposed approach uses both color and texture features for cloud detection. The input image is transformed into Lab color model for extracting the color features and gray scale image for extracting the texture features. Transformed images are converted into neutrosophic domain. An indeterminacy reduction operation is performed for getting better results. Finally a Fuzzy C-means clustering is performed on the true subsets. This gives the cloud detected image. This method is efficient in detecting thick clouds and thin clouds in Landsat images. Result analysis shows that the proposed algorithm can effectively detect the thin cloud. The proposed algorithm gives accurate results in less time complexity.
Highlights
When the world’s first satellite pictures of the atmosphere were viewed, the most remarkable feature was the extensive cloud cover over large parts of earth
We propose a cloud detection algorithm based on wavelet transform and neutrosophic sets [13,14]
Thick-cloud contamination is a common problem in Landsat images, which limits their utilities in various land surface studies
Summary
When the world’s first satellite pictures of the atmosphere were viewed, the most remarkable feature was the extensive cloud cover over large parts of earth. Persistent cloud covers over many regions cause difficulties in remote sensing with optical satellite imagery. Land scenes are, on average, about 35% cloud covered, as reported by Ju and Roy [1], indicating that cloud covers are usually present in optical satellite images. This phenomenon limits the usage of optical images and increases the difficulty in image analysis. Cloud detection is useful in improving the accuracy of land cover classification in multispectral images. Cloud cover is a big challenge in optical remote sensing of the earth surfaces, especially over the humid tropical regions.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
