Passive Multidimensional Imaging and Recognition with Multiple Degrees of Freedom

Abstract

The purpose of this project is to introduce, investigate, and demonstrate a new multi-dimensional sensing capability where a small number of photons may allow the possibility of implementing reliable multidimensional visualization, identification, and classification using polarimetric and multi-spectral band passive imaging. The multidimensional sensing and imaging proposed here are intended to operate under extremely sparse (low photon count) data conditions to enable many benefits, including operation in degraded environments, removing of occlusion to allow visualization of occluded objects, multi-dimensional target recognition, and tracking of occluded objects. Unlike LADAR, this passive multi-dimensional sensing approach uses 2D image sensors with ambient light or thermal imaging for persistent and covert operation. This capability will be very beneficial to enhancing surveillance, monitoring human activities, and information exploitation. During the course of this project, we have performed experiments and theoretical analysis of the objectives and tasks of the project. We have investigated polarimetric multidimensional imaging in degraded environments. We have investigated deep convolutional neural network model using physics based training with three-dimensional (3D) imaging to recognize polarimetric 3D objects in degraded environments such as low light and partial occlusions.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 18, 2022
Accession Number
AD1190034

Entities

People

  • Bahram Javidi

Organizations

  • University of Connecticut

Tags

Communities of Interest

  • Sensors

DTIC Thesaurus Topics

  • Accuracy
  • Algorithms
  • Artificial Intelligence Software
  • Computer Vision
  • Convolutional Neural Networks
  • Deep Learning
  • Detection
  • Detectors
  • Human-Machine Interaction
  • Image Recognition
  • Kidney Diseases
  • Materials
  • Neural Networks
  • Object Recognition
  • Optical Detection
  • Optics
  • Pattern Recognition
  • Polarizers
  • Target Recognition
  • Three Dimensional

Readers

  • Medical Imaging.
  • Neural Network Machine Learning.
  • Sensor Fusion and Tracking Systems.

Technology Areas

  • AI & ML