Spectral NLOS imaging- towards photo-realistic NLOS reconstructions

Abstract

In this 3 year project we will conduct the first in-depth study of hyperspectral Non-Line-of- Sight (NLOS) imaging using Single Photon Avalanche Diodes (SPAD) across the visible, near infrared (NIR) and Short Wave Infrared (SWIR) spectrum up to 1600 nm. We will create spectral NLOS data and develop several super-resolution and pattern recognition algorithms that specifically depend on integrating spectral information into the reconstruction process. We aim to create NLOS images that resemble a line-of-sight image of the scene seen rivaling the spatial resolution of the human eye, as well as demonstrate NLOS pattern recognition capabilities that match or outmatch human capabilities for specific tasks. Our overarching goal is to better understand on the information value of spectral data in NLOS imaging.

Document Details

Document Type
DoD Grant Award
Publication Date
Mar 07, 2023
Source ID
FA95502110341

Entities

People

  • Andreas Velten

Organizations

  • Air Force Office of Scientific Research
  • United States Air Force
  • University of Wisconsin System

Tags

Fields of Study

  • Physics

Readers

  • Computer Vision.
  • Sensor Fusion and Tracking Systems.
  • Spectroscopy.

Technology Areas

  • AI & ML