Automatic Detection of Type III Solar Radio Burst.

With the accuracy improvement of radio telescopes, massive amounts of solar radio spectrum data are received every day. It is inefficient to detect solar radio burst by astronomers, and it is also difficult to meet the real-time requirements of space weather, aerospace and navigation systems and etc. In order to reduce the workload of astronomers and improve the detection accuracy and efficiency, we propose an algorithm for automatic real-time detection of solar radio bursts based on density clustering in this paper. The algorithm firstly uses channel normalization to remove the interference of horizontal stripe in the image. Then, the normal distribution model is used for binarization, and then the DBSCAN clustering algorithm is used to cluster detection of the binarized solar radio burst area. Finally, the Canny operator is used to detect the edge and the time parameter of burst is extracted. Experiments show that the proposed method improves the detection efficiency and accuracy compared with some traditional clustering detection algorithms.