Hyperspectral Image Data Construction and Expansion Method of Ground Object

Hyperspectral image is an image that contains both geometric space information and spectral information, however, the cost of acquiring a large number of hyperspectral image data samples is very high due to the limitations of imaging technology and detector level.With the development of artificial intelligence technology and the advent of the era of big data, data sample build and extend the method arises at the historic moment. In order to obtain sufficient sample data and establish the hyperspectral data information database of specific ground object required for research, the obtained hyperspectral image data needs to be expanded which can provide basic data support for the study of spatial spectral characteristics of ground objects and subsequent hyperspectral image classification and target detection. In this article, Firstly, the data structure and characteristics of hyperspectral image are analyzed, and then the common methods and significance of hyperspectral image data sample expansion under the condition of land-based imaging are systematically described from the two aspects of spatial dimension data and spectral dimension data.Finally, Aiming at the main problems existing in the process of hyperspectral image data expansion, the development direction in the future is prospected.