Adaptive Sensing for High-throughput Imaging

Abstract

This project aims to assess the capabilities of capturing and processing very large-scale image datasets coupled to powerful computational backends. We will develop exible tools for design-ing maximally informative capture strategies for multi-dimensional object reconstruction. We will design and build high-throughput multi-sensor image acquisition systems coupled to optimization- based inverse algorithms. We will derive theory for algorithmic self-calibration, in which the hard- ware system is calibrated from image data, making reconstruction more robust and practical. Finally, we will explore ideas of adaptive system design for sparse measurements, as they pertain to large-scale acquisition. The resulting systems and theory that will be developed represent a new era of increased image data throughput, from gigascale/second towards terascale/second. The ideas and devices generated should have use in various naval applications including object tracking, surveillance and imaging in degraded environments.

Document Details

Document Type
DoD Grant Award
Publication Date
May 05, 2017
Source ID
N000141712401

Entities

People

  • Laura Waller

Organizations

  • Office of Naval Research
  • United States Navy
  • University of California Regents

Tags

Readers

  • Distributed Systems and Data Platform Development
  • Image Processing and Computer Vision.