Fast calculation of Adaptive-Non-Negative-Eigenvalue-Decomposition employing particle swarm Optimization

This paper considers a fast calculation of adaptive-non-negative eigenvalue-decomposition (ANNED). ANNED is one of useful polarimetric decomposition techniques and provides valuable information for remote sensing from measured covariance matrix. It has a disadvantage to take long time to estimate the decomposition parameters. To overcome this problem, the generalized eigenvalue problem has been applied. However, in practical use, more speed up is needed. In this paper, a particle swarm optimization (PSO) is applied to ANNED with the generalized eigenvalue problem. Thus, more fast calculation of ANNED is achieved and it is possible that POLSAR image with many pixels is analyzed.