Font Size: a A A

Research On Processing Technology Of Multi-Channel Satellite Cloud Images

Posted on:2018-11-16Degree:MasterType:Thesis
Country:ChinaCandidate:D N LiuFull Text:PDF
GTID:2428330518958900Subject:Electronics and Communications Engineering
Abstract/Summary:PDF Full Text Request
We can find various forms of cloud and a lot of weather information in the satellite images.According to the shapes and regions of the clouds,we can sum up the weather conditions and the possible occurrence of meteorological disasters.But in fact,the utilization of satellite imagery is very low.In order to make full use of multi-channel satellite clouds map,image fusion,mathematical morphology segmentation,and pseudo-color processing are used in the study of satellite images.(1)By integrating the data of the infrared cloud and the water vapor map,the information of the different channel satellite images can be better preserved,and the infrared cloud and the water vapor map can be preliminarily synthesized by using the non-subsampled contourlet transform(NSCT)method.Firstly,the NSCT decomposition operation is carried out on the water vapor map and the infrared image respectively.Then using the regional energy algorithm to fuse the low-frequency part and the region matching ratio algorithm to the high-frequency part.Finally,the fused image is reconstructed by NSCT inverse operation.(2)After the fusing the images,using two kinds of morphological process to segmentation the main cloud system.Firstly,I deal with the binary image by using the method that based on Otsu method and morphology.The Otsu method can select the appropriate threshold for the binarization operation on satellite cloud map.Then we use the mathematical morphology to select the appropriate structural elements and carry out a series of operations on the main cloud system to extract the main cloud system in the cloud map.Then the method of top hat transformation and Otsu method is also used to extract the main cloud system of cloud image.Comparing the experimental results with the threshold method and the method of regional segmentation,it can be concluded that the method adopted in this paper is more effective.(3)Finally,the extracted cloud map is pseudo-color processing,the thicker areas of the cloud can be enhanced.Through the subjective observation and the evaluation index analysis results,this method can keep the effective information of two-channel satellite cloud image,eliminate the irrelevant background,filter out the noise,keep most of the original cloud image and extract the main cloud system.The effect is good.
Keywords/Search Tags:satellite cloud picture, image fusion, mathematical morphology, image segmentation, pseudo color
PDF Full Text Request
Related items