Objective Center Location Method Research Of Tropical Cyclone Based On Infrared Satellite Image

A Tropical Cyclone(TC)is a strong weather system generated in tropical or subtropical surfaces.In the analysis and forecasting of TC,determining the real-time location of its center accurately is important.The precise location of TC is an important parameter of TC strength estimation.IR images from geostationary satellites are the most a reliable source for obtaining the location of TC,because of the high temporal sampling and wide coverage.The paper respectively proposes three tropical cyclone objective location method:Infrared brightness—temperature variance method for objective location of TCs,wavelet transform combined with sliency detection method for objective location of TCs and deviation angle gradient distribution uniformity(DAGDU)method for objective location of TCs.Then we can get the average location deviation contrast to the center position issued by China Meteorological Administration(CMA),the Japan Meteorological Agency(JMA),and the Joint Typhoon Warning Center(JTWC).The main work of this paper can be divided into three parts::(1)The objective location of TC based on the infrared brightness temperature variance method.First,an area of interest is selected from an infrared satellite cloud image.On the basis of this image,the histogram segmentation method is employed to obtain the binary image of the TC core region.K-means clustering segmentation is used to deal with the cloud region with violent changes in infrared bright temperature.The two binary images are combined to obtain a new binary image.The resulting binary image is detected by the Hough transform approach to reduce the search range of the center position of the TC.Finally,the deviation angle matrix is calculated from each pixel in the detection area as the reference center.The variance matrix is obtained by calculating the variance of the deviation angle matrix and filling it in the corresponding reference center pixel point of the detection area.The minimum value position of the variance matrix is taken as the corresponding TC center position.Then we can get the location deviation of the contrast to the center position issued by CMA,JMA and JTWC。(2)Wavelet transform combined with saliency detection method for objective location of TCs.The Bezier histogram and k-means clustering method were changed into the saliency detection segmentation algorithm and the wavelet transform image segmentation algorithm respectively on the basis of(1).And we changed the way to get the detection area.We used the saliency detection to get characteristic value and average position of the each superpixels,then we can get the new detection area.The results also show that,the average location deviation of the eye TC is less than the non-eye TCs,and the deviation of CMA and JMA is smaller than the deviation of JTWC.Compared with the previous method,the segmentation method of histogram segmentation combined with mean clustering is better than that of wavelet transform combined with sliency detection。(3)Deviation angle gradient distribution uniformity(DAGDU)method for objective location of TCs.Like the energy of Gray-level Co-occurrence Matrix,we change the deviation angle variance to the DAGDU without changing the segmentation method.This variable can reflect the distribution uniformity of the deviation angle matrix and can also reflect the thickness of the texture.Finally,the deviation angle matrix is calculated from each pixel in the detection area as the reference center.The DAGDU matrix is obtained by calculating the DAGDU of the deviation angle matrix.The maximun value position of the DAGDU matrix is taken as the corresponding TC center position.Compared with the first method,the DAGDU is better than deviation angle variance.And then three TC with different intensity are used to locating the center by this method.The results show that the stronger TC is,the smaller the locating deviation.