Progress and Application of Multi-Spectral Data Fusion Methods

Modern spectral technology has developed rapidly, which can directly carry out rapid and nondestructive analysis of liquids and solids. Different types of spectra have different analysis emphases. To achieve more comprehensive and accurate analysis, it needs to be combined with a variety of analysis methods. Multispectral data fusion technology is to optimize and integrate different types of spectra to realize the complementary advantages of single spectrum, so as to obtain more comprehensive, more reliable and richer characteristic data and improve the accuracy and stability of model prediction. This paper introduces the fusion strategies of low, medium and high levels data fusion by fusion structure, respectively, evaluates the advantages and disadvantages of each fusion strategy, summarizes the new strategies and methods of multispectral data fusion technology in recent years, and focuses on the research and application progress of this technology, The fusion strategies and methods of different spectral technologies in different fields are listed in this paper, and the application prospects of multispectral data fusion technology are discussed.