Advanced Reconnaissance Sensors 2021-2023 - Final report

Authors:

  • Johan Eriksson
  • Maria Axelsson
  • David Gustafsson
  • David Bergström
  • Johan Berneland
  • Anderas Brorsson
  • Carl Brännlund
  • Alexander Forsman
  • Hanna Hamrell
  • Ludwig Hollmann
  • Vera Håkansson
  • Britta Levin
  • Johan Rideg
  • Daniel Svedbrand
  • Thomas Svensson
  • Erik Valldor
  • Niclas Wadströmer

Publish date: 2024-02-26

Report number: FOI-R--5553--SE

Pages: 54

Written in: Swedish

Keywords:

  • reconnaissance
  • surveillance
  • electro-optical
  • infrared
  • Vis/NIR
  • SWIR
  • LWIR
  • night vision
  • polarimetric imaging
  • multispectral imaging
  • hyperspectral imaging
  • neuromorphic imaging
  • machine learning
  • deep learning
  • target detection
  • compressive imaging

Abstract

This report summarizes the results from the research project Advanced Reconnaissance Sensors during the three years 2021-2023. The work is performed within the Swedish Armed Forces research and development program for Sensors and low observables (FoT SoS, AT.9229423). Passive electro-optical technology for night-vision, hyperspectral imaging and polarimetric imaging is studied. These three fields has earlier been identified as interesting for both enabling improved capabilities in the future as well as being potential threats. A fourth topic, aimed at new sensor technology, focusses on neuromorphic imaging. Results from a large and diverse set of areas are presented in the report. Improved detection and reconnaissance from detailed analysis of surface chemistry and geometry thru sensor technology and signal processing for hyperspectral and polarimetric imaging, both used separately and also combined, are demonstrated. New digital night vision sensor technologies that enable autonomous processing of image information are demonstrated through multimodal image fusion and image denoising. Neuromorphic sensor technology, which offers large dynamic range and high temporal resolution compared to conventional imaging sensors, is assessed in a wide range of applications.