One-dimensional image recognition of radar target based on particle swarm classifier

A novel particle swarm classifier is designed for one-dimensional image recognition of radar target. After pretreating the data, the classification rules are discovered by particle swarm optimization algorithm based on the training samples, and then the one-dimensional image of radar target is recognized by the discovered rules. The convergence of the new algorithm based on Bayes's theorem and stochastic transform process are analyzed. Experimental results on the data of three airplanes obtained in a microwave anechoic chamber show that the proposed method has a stable performance in recognizing both the one-dimensional images without noise and the one-dimensional images with noise.