Research on supervised classification of fully polarimetric SAR image using BP neural network trained by PSO

Supervised classification of fully polarimetric SAR image using neural network is a common method nowadays. As an effective learning method of neural network, BP algorithm is the most widespread one in the neural network algorithms. However, BP network is easy to fall into local extremum and exists shortcomings such as the slow training process. To this end, this paper presents a method of supervised classification of fully polarimetric SAR image based on particle swarm optimization algorithm and BP algorithm. This method can improve BP algorithm using PSO and increase the convergence speed as well as the training accuracy of BP network. Experiment using fully polarimetric SAR image show that the supervised classification result of this method is better than the traditional BP algorithm classification result.