Hidden Markov Model Based on Target Narrow Pulse Laser Transient Characteristics

Abstract

Aiming at the fact that the narrow pulse laser transient characteristics of target have not been applied in target recognition, a hidden Markov model (HMM)based on transient characteristics is proposed. For the scattering characteristics of each target in different poses, the reliable model parameters of each target were completed by using the training samples, and each hidden Markov model was established. The maximum likelihood of the test samples for each model was calculated, the target feature class corresponding to the largest probability value was selected as the output category. The result shows that target recognition by hidden Markov model has better performance in terms of calculation speed and accuracy. This method is fast and effective.