The Research On The Recognition Of The Classification Of Cloud, Typhoon Segmentation And Locating Of The Typhoon's Centre Based On The Meteorological Satellite Cloud Image

The meteorological satellite remote sensing, developed since 1960s, is an important breakthrough of the meteorological sounding techniques. The continuity of the meteorological satellite cloud images provided by it in the time and space ,can't be compared with any one of the previous sounding means. Moreover,the meteorological satellite cloud images play a extremely important role in the making of the weather forecasting and the monitoring of the atmospheric environment.But the current traditional artificial visual interpretation and mapping is still one of the major means to analyze the the meteorological satellite cloud images,which not only includes some subjective factors to a certain extent,but also don't benefit to the complete extraction and maximum utilization of the meteorological satellite cloud images' abundant information and obstructs the scientific,automatic and quantitative development tendency of the making of the weather forecasting at the same time.Therefore, the automatic extraction,quantitative discrimination and the realizations by the computer about the meteorological information (such as the recognition of the classification of cloud, the typhoon segmentation and the locating of the typhoon's centre) in the meteorological satellite cloud images is the mainstream during the information processing of the meteorological satellite cloud images in the current world.The FY-2C meteorological satellite (the first Chinese operational stationary earth orbit meteorological satellite) launched in October , 2004, in China , observes the Chinese mainland and sea area from the most perfect stationary earth orbit position (105°E longitude over the equator) and sends back the cloud images with the large coverage and all-weather cloud information of forming, disappearing, aggregating and diverging.Therefore,it's the hot spot concerned by people in recent years to apply the domestic FY-2C meteorological satellite cloud images in order to explore on the recognition of the classification of cloud, the typhoon segmentation and the locating of the typhoon's centre.The paper based on introducing the basic instance of the FY-2C meteorological satellite and it's flow of the images' processing,and inquiring the theories and the common methods of applying the meteorological satellite cloud images to research on the recognition of the classification of cloud, the typhoon segmentation and the locating of the typhoon's centre,utilizes the software of MATLAB₇.0, has the original FY-2C Meteorological Satellite's Lambert-projected cloud images as the samples and makes the research in the following three major aspects:(1) Comprehensively using the multispectral threshold method, mathematical morphology techniques, texture analysis methods, etc, in light of the characters of the multi-channel FY-2C Meteorological Satellite's Lambert- projected cloud images,so as to make the simulation experiments on the automatic recognition of the cumulonimbus. The gained results are satisfying.(2) Comprehensively using the threshold method, mathematical morphology techniques, mathematical statistics methods, etc, in light of the typhoon's main galaxy with the characters of higher grayscale value, bigger area, limited acting region and concentrated pixels, so as to make the simulation experiments on the typhoon galaxy's segmentation. The gained effects are satisfying.(3) Comprehensively using the multispectra method, mathematical morphology techniques, texture analysis methods, etc, in light of the eye area of the eyed typhoon galaxy with the characters of relatively lower grayscale value, relatively stronger circle-like quality,unique texture features,so as to make the simulation experiments on the automatic locating of the eyed typhoon's centre.The gained effects are perfect.