Research Of Typhoon Image Fusion Based On Multi-scale Geometric Analysis And Partial Differential Equation

China is one of the most serious countries that bothered by typhoon. There are generally seven typhoons every year, which load in China. In summer and autumn, the typhoon is one of the most severe weather systems in the southeast coastal in China. It brings heavy winds, storm and surge disaster, and causes people's life, state property, industrial and agricultural production losing heavily. In the opposite, the rain caused by typhoon can relieve widespread drought, the accurate, timely typhoon forecast and warning is useful for us, and the accurate center location is the basis of successful typhoon forecast. The center position of typhoon has been being the main work of satellite cloud image research. The series of satellite cloud images of the FY spectrum contains five-channel satellite images:IR1,IR2,IR3,IR4 and Visible images.The different characteristics of typhoon cloud images are contained in different channels. If we can fuse the characteristic information of different channels, it will be helpful for meteorological department to research the characteristics of typhoon and prevent typhoon disaster. Further more, it can help entire country defense typhoon and reduce the damage caused by typhoon disaster. The main research works are as follows:(1) A new feature point extraction algorithm which is used to match the satellite cloud image feature point is proposed. The new algorithm can do well in registering two images by optimizing registration parameters of affine transformation function about the corner information of the curvature of images. The experimental results show that the proposed algorithm can accurately match the satellite cloud images and better than traditional image registration methods.(2) An effective method of multi-channel typhoon cloud image fusion is proposed by combining wavelet transform with partial differential equations (PDE). On the one hand, this method can do better for multi-channel typhoon image fusion, on the other hands, PDE is used in the fusion to reduce the noise. As the last work, the final fused image can contain the information of each channel images. Compared with the traditional methods of image fusion, the proposed method can improve the contrast of the fused typhoon cloud images largely.(3) An effective method of multi-channel typhoon cloud image fusion is proposed by combining contourlet transform with partial differential equations (PDE). A nonlinear enhancement operator is used to enhance the details of the high-frequency cloud images in contourlet domain. The proposed method can do well in keeping the important details and contour information of fused images. The noise can be reduced during the cloud images being fused. In short, the final fused image can contain much information, and improve the sharpness of the fused image.(4) A new method of multi-channel typhoon cloud image fusion is proposed by combining contourlet transform with particle swarm optimization. Actually, a function is constructed by combining entropy, gradient with signal to noise ratio of multi-channel typhoon images. Then, the function is optimized by particle swarm optimization, which can make the fused images containing more information. The experimental results show that the proposed method is better than some traditional image fusion methods.