Polarization Gradient Histogram for Object Tracking in Infrared Polarization Imaging: A Feature Extraction Method for Polarization Mosaic Image

Xinbo Qiao, Lulu Chen, Yongqiang Zhao

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

4 Scopus citations

Abstract

Infrared polarization feature improves the performance of anti-interference for object detection. Division-of-focal-plane polarization imagers (DPIs) make it possible for real-time object tracking in infrared polarization imaging. Output of DPIs is polarization mosaic image. Traditional processing way is combined by demosaicing and then computing polarization parameters for object detection and tracking, which is time consuming and subjected to demosaicing error. This paper proposes a histogram of infrared polarization mosaic gradient (HIPMG) for feature extraction in infrared polarization image. In our proposed algorithm, Polarization and Spatial Correlation Matrixes (PCM and SCM) are designed to obtain spatial and polarization correlation of the object. Based on PCM and SCM, two Polarization Filter Matrixes (PFM) are proposed to obtain spatial-polarization gradient maps (SPGMs). Then a HIPMG descriptor is designed from the intensity and direction of SPGMs. A polarization-track method is proposed by using HIPMG for object tracking. Experiment results demonstrate the effectiveness of proposed feature extraction method over the state-of-the-art methods.

Original languageEnglish
Title of host publicationICCPR 2021 - Proceedings of 2021 10th International Conference on Computing and Pattern Recognition
PublisherAssociation for Computing Machinery
Pages224-230
Number of pages7
ISBN (Electronic)9781450390439
DOIs
StatePublished - 15 Oct 2021
Event10th International Conference on Computing and Pattern Recognition, ICCPR 2021 - Virtual, Online, China
Duration: 15 Oct 202117 Oct 2021

Publication series

NameACM International Conference Proceeding Series

Conference

Conference10th International Conference on Computing and Pattern Recognition, ICCPR 2021
Country/TerritoryChina
CityVirtual, Online
Period15/10/2117/10/21

Keywords

  • Feature Extraction
  • Infrared Polarization
  • Object Tracking

Fingerprint

Dive into the research topics of 'Polarization Gradient Histogram for Object Tracking in Infrared Polarization Imaging: A Feature Extraction Method for Polarization Mosaic Image'. Together they form a unique fingerprint.

Cite this