Non-line-of-sight imaging using multi-bounce light
Name
1102050128-MIT.pdf
Size
6.57 MB
Format
Adobe PDF
Checksum (MD5)
4a16954f32d2c37c6fefe330c66d1e74
Author(s)
Henley, Connor A.
Advisor(s)
Franco N. C. Wong and Jeffrey H. Shapiro.
Date Issued
2019
Publisher
Massachusetts Institute of Technology
Abstract
Non-line-of-sight (NLOS) imaging techniques produce images from light that has travelled from the scene of interest to the observer via indirect paths which typically include multiple reflections. Such techniques can be particularly useful when the direct line of sight between the observer and the scene is blocked. In this thesis we will explore two NLOS imaging techniques. The first is an occlusion-assisted imaging technique, which constructs images of hidden scenes by interpreting the patterns that are imposed on multiply reflected light by occluding objects. We will provide a conceptual and theoretical introduction to our technique, which uses a focused, scannable illumination source and a single-pixel, lensless detector. We will then present the results from an experimental implementation of this technique in a challenging environment. This will be followed by an analysis of a number of challenges that are commonly encountered in active, occlusion-assisted imaging scenarios, including single-bounce light rejection, inter-reflections, and asymmetries in measurement geometry. Finally, we will introduce a new NLOS imaging technique which uses the time-of-flight information in multiply reflected light to produce an unobstructed, line-of-sight view of a hidden scene. We will provide a conceptual introduction to the technique as well as a derivation of the physical model that underlies it, and will also discuss methods for visualizing the technique's output.
Description
This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Thesis: S.M., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2019
Cataloged from student-submitted PDF version of thesis.
Includes bibliographical references (pages 91-93).
Subjects
Electrical Engineering and Computer Science.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission.
Persistent DSpace Link
