Speckle filtering in polarimetric SAR images based on improved independent component analysis
Liu Yu-d · Cehui kexue · 2015
The speckle filtering in polarimetric SAR images is very important to its post-processing.For that the independent component analysis(ICA)is unstable and easily falls into the local optimum,an improved independent component analysis based on particle swarm optimization(PSO)algorithm was proposed in this paper to reduce the polarimetric SAR image speckle.The algorithm focused on improving the generation process of the un-mixing matrix W by PSO algorithm,and improving the separation efficiency of the original ICA algorithm.In order to verify the validity of the improved ICA method,the Lband data of the San Francisco bay were tested,which were obtained by AIRSAR.The efficiency was evaluated by the speckle reduction index,iterations and convergence time.The experiment results showed that the improved ICA algorithm based on PSO had higher convergence speed and better speckle reducing efficiency.