Research on Optical Rainfall Calculation Method Based on FPGA

For the forecast and detection of disastrous precipitation weather, the most important thing is the detection of rainfall. However, the rainfall sensors commonly used at present, tipping bucket or weighing rain sensor, cannot meet the exist actual situation owing to the long-time of manual maintenance, measurement distance is short and the real-time performance is poor. The optical rain gauge has obvious advantages since its introduction. It can monitor the raindrop size and measure the speed of raindrops in real time. It has strong adaptability and can be free of maintenance for a long time. It can be applied to real-time monitoring of rainfall in harsh environments and remote areas. This paper mainly introduces the image algorithm processing on the raindrop image collected by the CCD camera. Initially use the acquisition card of the CCD camera to capture the original image data, and the image processing algorithm matching research is using the matlab. Then, after the algorithm obtained by the research is mapped to the FPGA, an automatic measurement, recognition, and discriminant rainfall model is generated by acquiring the image data collected by the line CCD and transmitting it to the FPGA. By taking advantage of the parallel processing of FPGA, the FPGA can quickly filter, binarize and shape the precipitation particle image data, and properly solve the problem of overlapping and occlusion between precipitation particles due to single plane acquisition, and finally deal with it. The image of a good precipitation particle is further calculated by the relationship between the plane equivalent diameter and its solid diameter, and the three-dimensional diameter of the precipitation particle is further calculated.