Research on Vectorization of Weather Radar Image

摘要

In this paper, a new image vectorization method of weather radar image has been put forward, based on the theory of raster data vectorization. First of all, according to the nature of weather radar image that displays by pseudo-color, the image is divided into some layers which are processed separately. Then, after the image binarization processed, the edge detection algorithm was exploited to extract the boundary of each binary layer. And the corner detection algorithm was applied to detecting the feature point set on the boundary. The cubic B-spline curve was used to fit the boundary. Finally, this algorithm can convert a weather radar raster image into a graphic. The experimental results showed that, this algorithm makes the radar echo chart the intensity boundary to be clearer, the shape to be clearer, the central value to be accurate, and the pattern recognition to be intuitive. It great enhanced the radar echo information distinction effectiveness and the rate of accuracy, further enhanced the accuracy of the radar echo information monitoring and the strong convection weather forecast. It also satisfied the application needs of thematic information mapping and database development.