Research On Image Fusion Arithmetic

The aim of image fusion is to combine different images from different sensors and get a more correct clear and complete knowing about the target than a single sensor.Application of multi-sensor image fusion span a broad range that includes military and civilian. Multi-sensor image fusion is one popular research field in image processing.This thesis explores and researches wavelets-based fusion method and intelligentized fusion method with the discussing of their theories and algorithms.Wavelets-based fusion method is a relative better method from mathematical theory to arithmetic model. In this part of research, at first, this thesis studies the fusion of two Panchromatic images.It makes use of pixel-based method and region-based method.Region-based method could improve the definition of image. At the same time, it corrects the fringe that using the pixel-based method produces.Then it researchs the fusion of Panchromatic and Multi-spectral image.The arithmetic exercises IHS transform and Wavelet Transform.Each image was decomposed into certain sub-images which are consisted of an approximate image and a set of spatially-oriented detail images by the tow dimension orthogonal wavelets. These approximate image and detail images were merged with different Fusion model, which could keep the spectral feature of the original multi-spectral image and improve the content of information effectively.Moreover, it has stronger self-adaptive activity.Neural-fuzzy is a part of Intelligentized fusion method, and is one of the methods that is at early stage of development, but has superperformance and great potential development. This thesis conbines IHS transform and neural-fuzzy for the sake of the fusion of Panchromatic and Multi-spectral image. The method based on the fuzzy logic attains one neural-fuzzy method by connecting to the neural network, uses both the learning capability of neural network and the decision-making capability of fuzzy logic. The training of membership functions is based on least square method and back propagation associated with gradient descent method. Finally the analysis and simulation are carried through.At the end of this thesis, it analyzes and compares these two fusion methods, and summarizes the advantages and disadvantages respectively.