Detection Of Hailstone In The Rainstorm

Hailstone is one of the main meteorological disasters, the automatic detection of the hailstone cloud been always the key point of research on the strong storm tracking. The accurate forecasting of the hailstone can notice the related department to do some management to prevent the hail, which can effectively reduce the damage of the hailstone.As the extensive use of doppler radars, image-rich products are obtained.But a great deal of high-dimensional data is brought. Actually, high-dimensional data are difficult to be used. Hence, data dimensiona reduction becomes very import. Rapidly changing IT technology can greatly improve our ability which to develop friendly interface software.The main work of the paper include:(1) The database of radar image feature is constructed. This will provide the data source for identification of Hail. It is an important part of Automatic Recognition of hail clouds.(2)Manifold learning and data mining algorithms are used for mining association rules. These rules are related to characteristics of radar images. Combined with meteorological expertise, useful knowledge are obtained.(3) With the rules which were obtained from the association rule mining of radar image,a new method of automatic recognition of hail has been achieved. Hailstorm in the radar image, can be highly efficient identification. Meanwhile, a new database development tool has been developed, and a novel interface display tools have been introduced.This study for the automatic recognition of hail clouds has provided a new research approach. This system meets the needs of the application.