Sparsity-motivated multi-scale histograms of oriented gradients feature for SRC

Suoqi Zhang, Jiulu Gong, Derong Chen*, Linfeng Xu, Lei Yan

*Corresponding author for this work

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

4 Citations (Scopus)

Abstract

In order to recognize targets accurately from the low-quality images obtained from unmanned system, sparse representation based classification (SRC) method using sparsity-motivated gradient feature was proposed. The multi-scale histograms of oriented gradients (HOG) feature was used as an original feature, whose dimension was reduced by a non-adaptive random projection method. A very sparse measurement matrix was adopted to preserve the structure of multi-scale HOG feature space efficiently. The sparse representation was obtained via i1-norm minimization, and the least reconstruction error was used as recognition principle. Experiment results again Comanche FLIR data set show that, the proposed method can raise the recognition rate by 2% compared with the state of art methods.

Original languageEnglish
Title of host publicationProceedings of 2017 IEEE International Conference on Unmanned Systems, ICUS 2017
EditorsXin Xu
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages389-393
Number of pages5
ISBN (Electronic)9781538631065
DOIs
Publication statusPublished - 2 Jul 2017
Event2017 IEEE International Conference on Unmanned Systems, ICUS 2017 - Beijing, China
Duration: 27 Oct 201729 Oct 2017

Publication series

NameProceedings of 2017 IEEE International Conference on Unmanned Systems, ICUS 2017
Volume2018-January

Conference

Conference2017 IEEE International Conference on Unmanned Systems, ICUS 2017
Country/TerritoryChina
CityBeijing
Period27/10/1729/10/17

Keywords

  • automatic target recognition
  • compressive sensing
  • histograms of oriented gradients
  • sparse representation

Fingerprint

Dive into the research topics of 'Sparsity-motivated multi-scale histograms of oriented gradients feature for SRC'. Together they form a unique fingerprint.

Cite this