Study On Image Fusion Based On Fast Fourier Transform

Image fusionis a technology that combines relevant information from two or more images into a single imageto obtain more accurate and comprehensive research objects.Image fusion technology has been applied to remote sensing astronomy,medical imaging,military security monitoring,computer vision and other important areas.Firstly,we propose an improved algorithm HIS + SWT.The step of the algorithm is: it use the method of color Intensity Saturation(Hue,Intensity,Saturation,HIS)to perform HIS transformation on the source image,which converts the image from RGB space component to the HIS.The source image is decomposed into four bands by stationary wavelet transform and then apply fusion fusion rules.Finally,the adjusted fusion image was obtained by inverse SWT.Compared with conventional methods,we can conclude that the algorithm proposed in this paper has more advantages and accords with experimental expectations.Secondly,for the Poisson fusion boundary unnatural problem,this paper proposes an improved Fast Fourier transform Poisson image fusion algorithm.The core step of the algorithm is to use the basic gradient decomposition method to decompose the image into the basic layer and details layer.At the base layer,using the improved fast Fourier transform Poisson image fusion algorithm to obtain a new base layer.The detail layer uses weighted combination to get a new detail layer.Finally the new detail layer and the base layer are merged into fusion image.Compared with the existing methods,the algorithm in this paper can solve the boundary unnatural problem better and show its superiority.