MidIR and LWIR Thermal Polarimetric Imaging Comparison using Receiver Operating Characteristic (ROC) Curve Analysis

Abstract

We report results from a field-test study to assess target detection capabilities for both mid-IR (MidIR) and long-wave IR (LWIR) polarimetric camera systems that was held during the week of January 19-23, 2020, at Fort Hunter Liggett, California. In particular, we analyze detection ability for a stationary target, over a prolonged period, in which receiver operating characteristic (ROC) curve analysis was applied to the resultant thermal and polarimetric imagery recorded by both sensors. Imagery considered includes conventional MidIR and LWIR thermal radiance, S0 images, as well as Stokes imagery (i.e., S1, S2), and degree-of-linear polarization (DoLP) for both wavebands. Both polarimetric imaging systems used Stirling cooled focal-plane arrays that were based on mercury cadmium telluride or indium antimonide substrates, with pixel densities of 640x480 and 1280x1024 for the LWIR and MidIR sensors, respectively. ROC curve analysis showed improved target detection for DoLP images for both wavebands when compared to conventional thermal imagery, S0 (W/(sr*m2)).

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Oct 01, 2020
Accession Number
AD1111272

Entities

People

  • Kristan P. Gurton
  • Richard Edmondson

Organizations

  • United States Army Research Laboratory

Tags

Communities of Interest

  • Advanced Electronics
  • Air Platforms
  • Sensors
  • Weapons Technologies

DTIC Thesaurus Topics

  • Abstracts
  • Antimonides
  • Arrays
  • California
  • Cloud Cover
  • Department Of Defense
  • Detection
  • Detectors
  • Field Tests
  • Focal Plane Arrays
  • Focal Planes
  • Improvised Explosive Devices
  • Indium Antimonides
  • Linear Polarization
  • Long-Wavelength Infrared Radiation
  • Measurement
  • Military Research
  • Polarimeters
  • Polarization
  • Radiance
  • Remote Sensing
  • Target Detection
  • Two Dimensional

Readers

  • Image Processing and Computer Vision.

Technology Areas

  • AI & ML