Research on Rainstorm Weather Detection Method Based on Weighted Density Adaptive A-DBSCAN Algorithm

Accurate detection of regional extreme weather is a research hotspot. Therefore, based on the real-time image of the camera, this paper proposes a new method for real-time detection of rainstorm weather based on the camera image by combining Laplacian operator and gaussian filter to calculate the image feature value (LOG value), and then establish the regression model of the LOG value and the recorded rainfall data. In addition, aiming at the characteristics of high noise in images collected by cameras, this paper proposes an adaptive method of A-DBSCAN denoising based on weighted density. The experiment in this paper proves that A-DBSCAN method can greatly improve the accuracy of real-time detection of rainstorm information from camera images.