A Method of Radar HRRP Aircraft Type Identification and Rejection Based on LSTM

Abstract

Aircraft type recognition refers to the identification of aircraft type by analyzing the characteristics of target echo data obtained from radar scanning. Aircraft type rejection refers to a judgment on the acquired high-resolution range profile data to determine whether the data belongs to the current model to be classified. In order to effectively use multiple feature fusion to improve the recognition performance, this paper proposes a model of aircraft type recognition and rejection based on LSTM. This model analyzes the HRRP features of aircraft through LSTM network, and classifies targets by full connection layer network. From the point of view of confidence of classification results, a method of aircraft type rejection based on softmax is proposed.