A Novel Measurement Matrix Optimization Approach for Hyperspectral Unmixing

Algorithm 1

The flow diagram of the proposed method. Optimization of TSCMM.

Input: number of measurements , dictionary , threshold

updating ratio , number of iterations

Initialization: Set to be TSCMM.

Update:       Set the initial value of iteration .

      Optimize Gram matrix

          (a) Compute Gram matrix: ;

          (b) Normalize: ;

          (c) Update the elements of Gram matrix ;

       Optimize measurement matrix

          (a) Set the initial value of iteration ;

          (b) Compute the orthogonal gradient factor matrix ;

           (c) Update measurement matrix: ;

           (d) , if , stop; else, return to Step  3.2);

      Compute measurement matrix ;

       , if , stop; else, return to Step  ().

Output: is the optimal measurement matrix .

Further, SNR and reconstructed signal.