Study On Revision And Calibration Method Of Optical Rainfall Sensor

Meteorological monitoring has been widely used in climate change research, disaster forecast, agriculture and other fields, at present is developing towards the direction of diversification and intensification. Traditional rainfall monitoring technology can not meet the requirements of modern meteorological measurements, optical rain sensor measuring system came into being, gradually become a hot research topic.In this paper, the error reasons of optical rainfall measuring system are conducted in-depth analysis.Then a calibration scheme is proposed, finally verified by experiments.Main contents are as follows:First, structure and working principle of two optical rainfall measuring system are introduced, comparing the advantages and disadvantages of the two systems, and this paper analyzes the causes of error, and proposed the corresponding solution.Secondly,image preprocessing:using gaussian filter for image denoising,using mathematical morphological method to smooth the boundary and fill the inner cavity,proposing a algorithm based on Hough transform to eliminate the dark grain of image background.Thirdly,depending on the degree of overlap of the image, adopt corresponding segmentation scheme.By analyzing the experimental data,a corresponding calibration scheme is proposed,and the feasibility of the scheme is verified by experiments.Finally,with raindrops simulation testing, area of each part of the corresponding image binarization threshold are collected, the calibration of measuring regional surface resulting from the analysis data. By experimental verification, calibration surface can significantly improve the measuring accuracy of optical sensing rainfall system.