Research On Compression Coding Of Weather Radar Echo Data Based On Prediction Algorithm

Weather radar can effectively monitor severe weather such as typhoon,rainstorm,hailstorm and tornado,and can quantitatively measure large-scale precipitation.In aviation meteorological support,weather radar plays an irreplaceable role in the detection of local convective weather.As an important means of weather detection,weather radar has been widely used in various countries.The new generation of Doppler weather radar can not only detect the position and intensity information of meteorological objects,but also obtain the radial wind speed and spectral width and other information of the radar,so a large amount of data will be generated in each detection.In the radar network detection business will produce massive data.Research on the compression and coding of weather radar echo data can save a lot of storage space and shorten data transmission time by reducing the data volume.These are of great significance to disaster prevention and reduction,national production,people's life and application in military field.For data compression coding,although there are a lot of research on neural network and fractal theory algorithms,in some still image and video compression industry specifications,the prediction coding,wavelet transform and statistical coding and other mature algorithms are still the main ones.Combining with a weather radar produced by Anhui sun create Co.,LTD.,this paper focuses on the research of lossless radar echo data compression coding based on prediction algorithm.In this paper,the storage mode of weather radar original data and the characteristics of echo image are analyzed.At the same time,several prediction algorithms and arithmetic coding principles are studied.Data preprocessing,predictive transformation and arithmetic coding are used to compress the radar echo data.During data preprocessing,the length of radar echo data is clipped,which effectively reduces the data volume.In order to make full use of the correlation between the data,the echo data are arranged in matrix form.LMS(Least Mean Square)prediction algorithm,improved block prediction algorithm and MED(Median Edge Detection)prediction algorithm principle were studied,and the compression performance of several algorithms was verified and compared with real echo data.The results show that the predicted residual entropy is the minimum after the transformation of radar echo data by MED prediction algorithm.In order to further improve the compression rate,the arithmetic coding of the first-order adaptive model is used to encode and output the predicted residuals.After a large number of data testing,the algorithm described in this paper is used to compress the weather radar echo data.The compression ratio can reach 17.5~55.6,achieving the expected goal of this paper.