基于数据同化的图像融合方法研究

数据同化凭借其在数值预报中将观测数据与理论模型相结合的特点，目前广泛应用于遥感图像处理及图像融合领域。文中介绍了数据同化的含义和数据同化系统的构成，并分析了其应用于图像融合的原理；结合现有的研究成果着重分析了数据同化框架下结合不同优化算法的图像融合方法，并对其进行分类比较，给出了各方法的优缺点和研究成果及应用；最后在总结了各种基于数据同化的图像融合方法普遍存在的问题的基础上，探讨了进一步发展与研究的方向。

Research on Image Fusion Method Based on Data Assimilation

Data assimilation with its benefits of combing observation data with prediction model in the field of numerical prediction is widely used in the field of remote sensing image processing and the image fusion. In this paper,the meaning of data assimilation and the structure of its system are presented,the principle of its applying into image fusion is analyzed. Referring to the existed research,also have a detailed analysis of different image fusion methods based on data assimilation and different optimization algorithms,further more,the work of classification and comparison is done,both the advantages and the disadvantages are emphasized and the direction of future re-search are showed. Then by summarizing the insufficient cased by the image fusion methods based on data assimilation,the development and the research direction of this method are discussed further.