Bionic Vision-based Multi-source Image Fusion Mechanism

Image fusion is an integrated information processing technology, which mainly aims to improve the reliability of images by processing the redundant information and the definition of images by processing complementary information among several images. What’s more, it is possible to fuse the complementary information to obtain a composite image with more detailed and complete information content. It has been widely used in medical imaging, remote monitoring, feature recognition, high dynamic range imaging, and image compression etc. This paper is mainly to discuss that visual attention mechanism is applied to wavelet image fusion.This paper firstly introduces the development of image fusion technology, basic theory, quality evaluation methods of image fusion, and then introduces the pixel level image fusion based on spatial domain and several image fusion algorithms based on frequency domain. It puts forward a multi-focus image fusion method based on edge enhancement, wavelet multi-scale decomposition of the two images to fusion, according to the maximum absolute value method for the fusion of high frequency components of image, approximate component through the fusion of edge information to guide the low-frequency intensity enhancement and extraction of the original image, in order to highlight the fusion image edge information in the. And the use of information entropy, standard deviation evaluation index to evaluate the method. The results show that the method has good fusion results, highlight the frequency domain of the edge detail information better, improve the fusion image clarity, improved visual effect.Due to considering the influence of the human eye’s visual features of image fusion technology, this paper mainly introduces the characteristics of visual attention mechanism, the basis theory of visual attention mechanism mode and calculation process. A new bio-inspired image fusion algorithm is proposed in this paper. Based on visual attention mechanism(VAM), visual saliency map is used to guide the fusion process of wavelet transformation coefficients. The main advantage of the proposed algorithm lies in using visual saliency maps to emphasize importance of interesting regions. visual similarity, a new fusion quality evaluating indicator which based on visual saliency maps comparison, is proposed to evaluate the algorithm performance. Experimental results demonstrate the superiority of the proposed algorithm in terms of subjective visual consistency and objective comparison.