Segmentation Method Based On Information Entropy Satellite Imageries Typhoon

The typhoon is near the center of the largest wind power reached12above the level ofthe tropical cyclone. The typhoon brought winds, torrential rain and storm surge, composedof a complex weather system, is the world of human production and life affect one of themost serious natural disasters. When the typhoon came by sea waves, after landing to bringstorm rainfall destroy crops, building facilities, on China’s workers and peasants, fishing,transportation and people’s property safety caused tremendous loss. Therefore, timely andaccurate prediction of typhoon, typhoon disaster losses can be reduced, to the social life, thestate property and the personal safety can have a profound effect. However, the prediction oftyphoon, typhoon segmentation and positioning and identification key, accurate and effectivemethod of typhoon segmentation not only on Typhoon positioning, recognition plays a crucialrole, also on typhoon forecast automation and atmospheric science research has importantsignificance.Information entropy is a measure of the probability of uncertain events, applied to imageprocessing is the measurement of gray level distribution random a characteristic parameter.Based on information entropy image segmentation method, can be in the segmentationprocess as much as possible to reduce the loss of image information, thus can be used incomplicated background segmentation. This paper used the satellite cloud is China is usingthe FY two meteorological satellite photographs transmitted back to the original image, andin2011the landing typhoon second " SED " and No. ninth typhoon " PLUM " as an example,based on the information entropy of the meteorological satellite cloud typhoon segmentationmethod to do a thorough research, mainly divides into the following several aspects:This paper analyzes the Typhoon Cloud formation, the process of development and thedemise, characteristics of each stage in the satellite cloud, summed research image processingapplications based on the satellite imageries split method and entropy theory at this stage.Describe the content and analysis of satellite cloud, Analysis of the basic characteristics ofsatellite cloud classification and cloud images, to compare Typhoon Cloud focus thecharacteristics of various types of cloud images, give each cloud type image characteristics inthe satellite and identification criterion.Study on Typhoon Cloud automatic segmentation method based on information entropy.First, this paper studied based on information entropy image segmentation techniques,analysis information entropy concept to the transformation of the image entropy. And demonstrated sensitivity angle from the mathematical properties of entropy and entropyinformation entropy is applied to the satellite cloud segmentation of effective and reasonable.Combined with the characteristics of the Typhoon Cloud satellite cloud image segmentationmethod based on information entropy, the area of the filter, the typhoon regional locationsegmentation, texture statistical properties of a typhoon automatic segmentation methoddesign. Experiments analyzed the effect of the segmentation and recognition.Study on Typhoon Cloud body segmentation method based on information entropy.Combined Two-dimensional maximum entropy method and two-dimensional Otsu, use multi-objective planning theory, fuse the two algorithms on a two-dimensional histogram, propose asearch method based on two-dimensional histogram quickly optimize. It adaptive adjustssearch, can effectively improve the operating speed of the two segmentation algorithm, andexperiments and analysis specific instances, compares segmentation results.