With the development of computer technology, the application of image processing is more and more extensive, in order to better use the radar cloud image to recognize the weather, this MATLAB program is used to denoising of the Radar cloud image from the median filter, and then 5*5 cellular neural network is used to extract the texture of the image, this article extracts energy, entropy, moment of inertia, correlation and deficit of the images by the gray level co-occurrence matrix method, the one, two, three order color moments and Hus InvariantM of the image as the feature data. Tectonic discrimination model utilizes AdaBoost classifier method to analyze Radar cloud image. Results show that the method can obviously distinguish Radar cloud image and the discriminant error is slight that has the reference value.
展开▼
