Approximate Sparse Regularized Hyperspectral Unmixing

Algorithm 1

Pseudocode of ASU algorithm.

Algorithm’s inputs: The measured hyperspectral data , spectral library , regularization , , parameter , the max iteration

number , and the iteration stopping criterion .

Algorithm’s output: The estimated abundance vector .

Initializations: Initialize , , and .

Main iteration as follows:

  Step  1. Compute by solving using (15).

  Step  2. Compute by solving using (18).

  Step  3. Update Lagrange multipliers according to (13).

  Step  4. Increase and go to step 1.

 Termination criteria: if or , stop iteration.