Study Of Image Fused Algorithms Based On Fractional Fourier Transform

Image fusion is a technology of synthesizing two or more images from different sensors or different conditions into a new image which can describes an object more accurate and more comprehensive. A lot of image fusion algorithms have been proposed in recent years, but they are used in image fusion confined to the time-frequency domain. Fractional Fourier Transform(FRFT) is a generalization of Fourier transform(FT), and has a more powerful ability of time-frequency analysis. Therefore fractional Fourier transform has a great research value in image fusion area.FRFT can present a signal in an arbitrary fractional fourier domain between time domain and frequency domain, and has a powerful flexibility. However, FRFT cannot obtain local information of a signal. Wavelet packet transform(WPT) is considered that a signal passes different scaled bandpass filters and has a strong ability of time frequency localization analysis. When FRFT is extended to FRWPT, FRWPT can decompose a signal in the fractional Fourier domain, and has the multi-resolution and fractional domain analysis characteristics. In addition, give the image fusion flow based on FRWPT. The fusion rule chooses the larger pixel value in the same position of different images. In order to reduce the noise and inconsistencies in fusion image, consistency checking method is added into the fusion process. Finally, the FRWPT fusion algorithm gets the better performance in the amount of information and definition than the mehods based on WPT, FRFT on visible and infrared images.After applying FRFT on an image, the reconstructed phase image can amplify high frequency components from the original image. Generally, the phase portion is of more importance. To take advantage of these characters, the phase and amplitude portions are fused separately. A novel image fusion algorithm, named as FRFT-phase-amplitude, is proposed. The experimental results indicate that the proposed fusion algorithm FRFT-phase-amplitude can improve the detail and definition in fusion images compared to these mehtods based on WT and FRFT.