Radar Target Detection Method Based on Neural Network Ensemble

Authors:

Abstract
The detection of small target in radar echo has important application value in military and civil fields. In the actual detection scene, the deep learning method is used to represent the complex clutter and targets, so as to realize the detection of small targets. However, the quality of training samples seriously affects the detection results. In order to make full use of radar echo information and reduce the influence of training samples on detection effect, a radar target detection method based on neural network ensemble is proposed. Firstly, the amplitude, power spectrum and time-frequency diagram are used as input to train three sub networks. Then, the three sub networks are integrated by using the stacking method to realize the target detection function. The test results of IPIX radar data show that this method can weaken the influence of the selection of training samples on the target detection effect and improve the detection ability.
Funding Information
  • National Natural Science Foundation of China (61871472)