Research On Image Fusion Algorithm Based On Multiscale Decomposition

As an important branch of information fusion technology,Image Fusion is a hot topic in current information fusion research.Image fusion combines multiple image information of the same target collected from different sensors,so that the image can contain more and richer information,and the expression is more accurate,thereby improving image quality.At present,image fusion technology plays an important role in civil,military,national defense and other fields.Therefore,research on image fusion technology is of great significance.Based on the multi-scale decomposition,this paper studies the multi-focus image fusion method and the remote sensing image fusion method.The main contents are as follows:(1)The background and significance of this paper are introduced,and the technical overview of image fusion,the classical method of image fusion and the evaluation system are expounded.(2)Multi-focus image fusion.In order to further improve the sharpness of the fused image while effectively retaining the edge contour of the image,a multi-focus image fusion method based on guided filtering and saliency multi-scale decomposition is proposed.Firstly,the wavelet transform is used to multi-scale the source image to obtain the low-frequency part and the high-frequency part of the image respectively.Then,the fusion method of the guided filtering is adopted for the low-frequency part,and the method of introducing the significant detection by the high-frequency part is merged.Finally,the fusion coefficients are wavelet reconstructed to obtain the fusion result.The simulation results show that the proposed algorithm has good fusion effect.(3)Remote sensing image fusion.In order to effectively solve the problem of fused image boundary blur and distortion,an remote sensing image fusion method based on IHS and saliency multiscale decomposition is proposed.Firstly,the multispectral image is transformed into IHS and the I component is extracted,the I component and panchromatic image are decomposed in multiple scales to obtain their high frequency information and low frequency information.Secondly,the classic weighted average fusion rule are used to fuse the low frequency information,while the fusion of the high frequency information uses the method of saliency detection.Finally,the fusion coefficients are multiscale reconstructed and IHS inverse transformed to obtain the fusion result.Experiments show that the proposed algorithm achieves better fusion results.