Research On Pixel Level Image Fusion Technology

Image fusion is evolved by information fusion, which enhances the reliability of fused image and gives a more objective, more accurate and more comprehensive description of the certain image scene through the processing of redundant information and complementary information of the source image. In recent years, image fusion has become a very important research field of many subjects and has been widely used. This paper mainly probes into the pixel level image fusion technology.This paper mainly does researches on the following:First, the paper describes a model of image fusion and image fusion level; it also studies the standards of subjective and objective quality evaluation of image fusion. It briefly introduces wavelet transform theory and the wavelet decomposition of the image. It also does researches on the image enhancement and edge detection before the image fusion takes place.Secondly, it studies image registration method based on the maximum mutual information. An algorithm which combines wavelet transform with maximum mutual information is put forward in this paper. The experimental results show that the method is effective and the effect of image registration based on maximum mutual information of second level wavelet transform is the best.Thirdly, compared with the classical image fusion method the paper points out the disadvantages of this new method. Based on the advantages and disadvantages of IHS transform and PCA transform integration the paper presents an algorithm that is a combination of the classical method and the new one. The experimental data shows that the algorithm gets better results.Finally, based on a comprehensive analysis of wavelet decomposition transform method of image fusion the paper proposes the wavelet decomposition based on regional integration rules. The original image is decomposed into low frequency and high frequency part through wavelet decomposition. For the low-frequency part, the weighting coefficient is used and for the high-frequency part and the components in different directions, different fusion rules are used according to the distinction of local energy. Simulation experiments are carried out by using visible light images, infrared images and left-focus images and right-focus images. The experimental results are obtained and the feasibility of the method is verified through an objective evaluation index. The analysis of experiment data shows that this method is more effective in image fusion compared with the method of wavelet transform based on the variation of pixel.