Study on Ballistic Missile Infrared Hyperspectral Endmember Extraction Algorithm Based on Non-negative Matrix Factorization——Performance Analysis

According to the hyperspectral endmember extraction algorithm based on the non-negative matrix factorization, the performance of this method in various situation is researched. The emphases of this research are the number of the hybrid spectrums, the abundance coefficient setting methods of the hybrid spectrums and the noise in the hybrid spectrums. The simulation show that the algorithm has good performance in various situations and this method can be used in the endmember extraction of the ballistic missile hyperspectral data.