Radar Emitter Recognition Based on SIFT Position and Scale Features

Shaokun Liu, Xiaopeng Yan*, Ping Li, Xinhong Hao, Ke Wang

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

59 Citations (Scopus)

Abstract

As the complexity of the battlefield electromagnetic environment has increased, it has become challenging to identify radiation sources using traditional radar emitter identification technology in low signal-to-noise ratio (SNR) environments. We propose a method of radar emitter recognition based on the scale-invariant feature transform (SIFT) position and scale features. The proposed method first uses time-frequency analysis to analyze the radar emitter signal, and the obtained 3-D time-frequency image is normalized into a 2-D gray image. Next, the Gaussian difference pyramid of the gray image is established, and the SIFT scale features and position features of the image are extracted based on the Gaussian difference pyramid. The extracted noise feature points are suppressed based on the scale features. Finally, a support vector machine is used for the automatic identification of radiation sources based on the SIFT position features. The simulation results show that the proposed method performs well in a low-SNR environment.

Original languageEnglish
Article number8326561
Pages (from-to)2062-2066
Number of pages5
JournalIEEE Transactions on Circuits and Systems II: Express Briefs
Volume65
Issue number12
DOIs
Publication statusPublished - Dec 2018

Keywords

  • Adaptive signal detection
  • image analysis
  • radar detection

Fingerprint

Dive into the research topics of 'Radar Emitter Recognition Based on SIFT Position and Scale Features'. Together they form a unique fingerprint.

Cite this