Detection and identification of chemical warfare and hazmat in the atmosphere using thermal hyperspectral and spectropolarimetric imaging

Overview
Project Logo

Context

Remote sensing detection and measurement of Chemical, Biological Nuclear or Radiological (CBRN) agents and hazmat (i.e. HazMat; hazardous chemicals) is becoming more important and is receiving much greater emphasis within military and security communities. The choice of the detection technique depends on the agent being sought, expected background interferences, and the required ranges. Most species can be detected and quantified due to their unique spectral properties in the thermal InfraRed wavelength regions. With tens to hundreds of spectral bands, thermal hyperspectral sensors pose an advantage over traditional open-path sensors in their ability to detect and quantify chemical molecular absorption and concentration along a large area of interest. Novel spectropolarimetric imaging provides complementary polarimetric information in large spectral dimension and therefore allows improving confidence in chemical identification and reducing false alarms rate. These pioneer technical combinations expect to outperform other spectral methods.

Objectives

  • Establishment of phenomenology knowledge behind the detection and identification of chemical precursors in the thermal hyperspectral (THI) and spectropolarimetric scenes;
  • Collection of a large set of advanced thermal hyperspectral and spectropolarimetric imaging data for the use of the Defence, academy and industry.
  • Development of unmixing models that are dedicated to thermal hyperspectral and specifically for chemical end-members extraction;
  • Development of robust methods for the detection of chemical warfare agents and hazmat spills using THI and spectropolarimetric data.

Methodology

To be able to understand the phenomenology behind the measurements in the thermal hyperspectral and spectropolarimetric scenes, the following methodology is implemented:

  • Development of physics based models that aggregate all the parameters in the scene and their combined influence on the retrieved radiance, polarimetry, emissivity and temperature.
  • Collection of extensive experimental data with numerous hyperspectral and spectropolarimetry imagers, in-situ ground sensors and ground truth data. During the trials, realistic scenes will be set and chemicals of interest will be released under controlled background and atmospheric conditions.
  • Validation of existing detection algorithms using real hyperspectral and spectropolarimetric data.
  • Development of robust methods for the detection and identification of chemicals precursors and spills in the atmosphere.
  • Creation of a road map for future technological development of hyperspectral and spectropolarimetric imaging.