Object Detection with Multi-band Polarization Imaging

Yongqiang Zhao, Chen Yi, Quan Pan, Yongmei Cheng, Seong G. Kong

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

5 Scopus citations

Abstract

Multi-band polarization imaging provides complementary information to conventional imaging sensors, such as material components and surface characteristics. An image data in multi-band polarization imaging makes an array of 3D data cubes. Each pixel corresponds to a certain point in a high dimensional space where a statistical analysis of the data leads to object detection. There are two aspects in object detection using multi-band polarization imaging. From a theoretical background of multivariate statistical analysis, multi-band polarization image data can be considered as common multivariate data to express statistical differences between the target object and the background. Physical properties of multi-band polarization imagery can be replaced by statistical differences to make the traditional optical remote sensing processing methods useful.

Original languageEnglish
Title of host publicationAdvances in Computer Vision and Pattern Recognition
PublisherSpringer Science and Business Media Deutschland GmbH
Pages111-153
Number of pages43
DOIs
StatePublished - 2016

Publication series

NameAdvances in Computer Vision and Pattern Recognition
ISSN (Print)2191-6586
ISSN (Electronic)2191-6594

Keywords

  • Adaptive Kernel Density Estimation
  • Anomaly Detection Method
  • Binary Hypothesis Testing
  • Conventional Image Sensors
  • Generalized Likelihood Ratio Test

Fingerprint

Dive into the research topics of 'Object Detection with Multi-band Polarization Imaging'. Together they form a unique fingerprint.

Cite this