Research On Pixel-level Image Fusion Method And Its Application

Image Fusion Technology, as an important part of the multi-sensor information fusion—the fusion of visual information, is a very popular research area in recent years. Image fusion is the technique that integrates and processes the multi-source images from various sensors intelligently and obtains more accurate, complete and dependable description about the same scene than any of the individual source images. The purpose of image fusion is to utilize the complementary and redundant information in multi-source images. As a result of this processing, the fused image should be more suitable for human visual perception or computer further processing tasks such as object recognition and feature extraction, etc.The pixel-level image fusion combines multi-source information at the basic level, and it can provide more abundant, accurate and reliable detail information than the fusion at feature-level or decision-level. So the fused images at pixel-level fusion contains more useful information that is required in further image analysis, processing and interpretation. The pixel-level image fusion technique is the most complex and difficult during the whole image information fusion. In this thesis, most of the research works focus on the pixel-level image fusion including the various fusion methods and their implementation, the evaluation of image fusion performance, the preprocessing technique during the fusion and the simple application of the pixel-level image fusion.This dissertation concentrated on the research works listed below and achieved some creative results.(1) Based on good understanding of the fundamental theory of image fusion technology and considering the disadvantage of the traditional fusion schemes which their multi-resolution representations are not suitable for the images that contain lots of important features in high frequency information, a new image fusion approach based on the discrete wavelet packet transform is presented for combining multi-source images at pixel-level. This algorithm performs the fusion process finely for both high frequency and low frequency parts of source images. Experimental results show that the proposed method can merge information from source images adequately and improve abilities of information analysis and feature extraction.. Then, for further improving the fusion effect, a new fusion rule based on region feature selection is proposed to make a fusion decision of wavelet packet coefficients. This...