Pixel-level Multi-sensor Image Fusion

Image fusion is one of the important parts of multi-sensor information fusion. Different image sensor can obtain redundant and complementary information. Image fusion combines all those different information to acquire an image that is satisfied with some special application. It aims at getting more exact and more comprehensive image description of the same sense or object by integrating the information of the multi-source images. Image fusion has become an active research topic with application in medical imaging, remote-sensing, computer vision and military. With the development of image sensor, people can obtain a great different image, but each image sensor has deficiency. So image fusion is a useful technology.Image fusion has three level, pixel level, feature level and character level. Pixel level is the elementary level, and it is the base of feature level and character level. This dissertation mainly aims at the research of multisensor image fusion algorithms in pixel level. First, the paper introduced the background of research, basic concepts, development; application of image fusion technology. The paper also elaborates the fundamental principles and methods employed in image fusion and objective evaluation criteria for image fusion performance. Second, the paper focuses on the wavelet transform based on image fusion. We introduced the basic theories, Mallat and á trous algorithms, and image based on wavelet. Then the paper proposed an image fusion algorithm using edge detection based on á trous algorithms. Third, the...