Classification Of Remote Sensing Image Based On Cloud Theory

With the development of space technology, remote sensing technology hascurrently become a kind of powerful technology to study earth space environment. Inremote sensing technology study, it dependents on the establishment of the remotesensing image classification of remote sensing information extraction, dynamic changeprediction, thematic map production and high resolution remote sensing database.Classification is the hot research contents of remote sensing technology. Although thereare a lot of methods of remote sensing image classification, but it can only solve fuzzyuncertainty or random uncertainty of the remote sensing image classification. Therefore,this paper introduced a method based on cloud theory. The main contents of this paperwere listed as follows:○1This paper introduced the basic concept of cloud theory. Mainly discusses thebasic definition of the cloud, cloud generator and virtual cloud. Cloud generatorincludes positive cloud generator, reverse cloud generator and condition cloud generator.At last, the paper introduces virtual cloud which uses the most times in this paper.○2Introduced classification method of remote sensing image based on cloud theory.According to the difference of cloud model,this paper introduced three classificationmethod of remote sensing image based on cloud theory.They are classification methodof remote sensing image based on the Normal cloud,floating clouds and comment cloud.○3Introduced classification experiment of remote sensing image based on cloudtheory. Compared with the traditional ISODATA and maximum likelihood classification,Simulation results show that cloud theory method has higher classification accuracy. Itis simple to calculate and less strict to equipment advantages.