APPROACHES TO AUTOMATIC DATA PROCESSING IN HYPERSPECTRAL REMOTE SENSING

摘要

In hyperspectral imaging, one has to deal with huge amount of data at different stages of their processing both on-board and on-land. Due to this, high degree of automation is required. The paper deals with considering new approaches and trends in design and use of methods and algorithms for automatic processing of hyperspectral data. Blind estimation of noise parameters is the first step. Then the obtained estimates are to be used at stages of image filtering and compression. Several strategies for doing this are considered. Their advantages and drawbacks are discussed. Examples of real-life data processing are presented.