Passive Multidimensional Imaging and Recognition with Multiple Degrees of Freedom
Abstract
Multidimensional Optical Sensing and Imaging Systems (MOSIS) will be investigated for object recognition and inspection in degraded environments, including obscuration, foliage, fog, sand storms, and photon starved scenes. The proposed approach is based on multi-view 3D photon counting integral imaging to record parallax and to extract range and 3D profile of targets. The proposed novel, transformative, multidimensional passive imaging system is capable of sensing, reconstructing, and recognizing objects and complex scenes from short range (meters) to long range (kilometers) using multiple degrees of freedom, including temporal, polarization, photon flux, and multispectral band information obtained from single or distributed image sensors on planar or non-planar surfaces. The proposed work is unique in having these degrees of freedom integrated in a passive imaging system for applications in degraded environments. The ability to detect, identify, and visualize targets and scenes passively in 3D including polarization and spectral data with only a few photons will be important for applications of interest to USAF, including space situational awareness, air-to-air and air-to-ground targeting, and small UAVs. Unlike LADAR, which uses active illumination to measure time of flight, this multi-dimensional passive sensing approach uses 2D image sensors with ambient light or thermal imaging. Photon counting detection reduces data by orders of magnitude through quantum imaging means. Thus, this approach provides high-speed computation for realtime processing of large data arrays, such as dynamic 3D plus spectral and polarimetric data, and reduces by orders-of-magnitude requirements on image data processing throughput and communications. Fundamental questions will be addressed for reliable performance-driven sensing, and performance evaluation of the multi-dimensional reconstructed scenes and object recognition will be conducted.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jul 11, 2018
- Source ID
- FA95501810338
Entities
People
- Bahram Javidi
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of Connecticut