A Study of Hailstone Detection System Based on Radar Echo Reflectivity Image

Abstract

Hailstone is one of the main meteorological disasters and it is very difficult to be forecasted effectively. A system of automatic recognition hail cloud is built base on the reflectivity image in this paper. The system deals the inputting image with the cut-down, filter, dilation and erosion process, and then the cloud is extracted from the image. A new concept called reflectivity density, which is used to distinguish hail cloud from rainstorm cloud, is defined. When recognized hailstone cloud the system will find the hailstone landing position by using Hough transform to detect hook echo or detecting the pixel with a reflectivity over 60 dBZ. This system is built on the Windows platform, and the experimental results validate the availability of the algorithm