Polarimetric HRRP Target Recognition Based on Convlstm

Wei Chen, Liang Zhang, Ying Xi, Yanhua Wang*, Yang Li

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

8 Citations (Scopus)

Abstract

The high resolution range profiles (HRRPs) of different polarimetric channels can enhance the recognition performance. Few existing methods focus on combining the structure information with correlation between different polarimetric channels. In this paper, we applied ConvLSTM to polarimetric HRRP target recognition. ConvLSTM can be used to combine the structure information with correlation due to the convolution operation and feedback mechanism. The HRRPs from different channels are regarded as the sequential inputs in ConvLSTM. The convolution operation can extract structure information and the feedback mechanism can exploit the correlation between different polarimetric channels. The experimental results demonstrate that the polarimetric HRRP recognition method based on ConvLSTM outperforms the state-of-art methods as well.

Original languageEnglish
Title of host publication2019 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2019 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages334-337
Number of pages4
ISBN (Electronic)9781538691540
DOIs
Publication statusPublished - 1 Jul 2019
Event39th IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2019 - Yokohama, Japan
Duration: 28 Jul 20192 Aug 2019

Publication series

NameInternational Geoscience and Remote Sensing Symposium (IGARSS)

Conference

Conference39th IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2019
Country/TerritoryJapan
CityYokohama
Period28/07/192/08/19

Keywords

  • convolution long short term memory (ConvLSTM)
  • high resolution range profile (HRRP)
  • polarimetric radar
  • radar automatic target recognition (RATR)

Fingerprint

Dive into the research topics of 'Polarimetric HRRP Target Recognition Based on Convlstm'. Together they form a unique fingerprint.

Cite this