Multiparameter Optimization for Mineral Mapping Using Hyperspectral Imagery

作者：Li, Na^*; Huang, Xinchen; Zhao, Huijie; Qiu, Xianfei; Geng, Ruonan; Jia, Xiuping; Wang, Daming

来源：IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018, 11(4): 1348-1357.

DOI：10.1109/JSTARS.2018.2814617

• Hyperspectral imaging system
• mineralmapping
• optimization modeling
• multivariate regression analysis

A multi parameter optimization model is proposed for the mineral mapping of hyperspectral imagery. This model can provide guidance for selecting system parameters when a new sensor is designed or offer performance estimation in mineral mapping with a given imaging system. A multivariate regression analysis is performed to investigate the quantitative relationship between mineral identification capability and imaging spectrometer parameters. The objective function of the proposed optimization model is set to maximize performance with the constraint on system parameters due to user requirements and technology maturity. The number of minerals identified is selected to measure the data usefulness. The parameters of an imaging system include ground sample distance, signal-to-noise ratio, modulation transfer function, and spectral resolution. A set of evaluation experiments is conducted using hyperspectral data collected in Dongtianshan area, Xinjiang, China. The predicted error is less than 15%, and the best mineral mapping can be obtained using the optimized system parameters.

• 出版日期2018-4
• 单位北京航空航天大学