Cloud detection and classification with artificial intelligence for satellite optical image

Abstract

Cloud is a very common weather phenomenon in the world. For optical satellite imaging, it is very often that more than 50% of imaging areas are covered by clouds. It sounds easy but practically very challenging to detect clouds accurately without confusing the detection with white ground, including areas covered snow or ice. This project is to study and test innovative approaches for accurate detection and classification of clouds by applying machine intelligence and big data. The project scope includes the study of image processing fundamentals, literature review of machine intelligence and cloud detection, implementation of the proposed approach, collection of cloud and snow samples, test the implemented code and detailed analysis of the results, evaluation of the detection accuracy.