Date of Completion

8-24-2018

Embargo Period

9-1-2018

Keywords

Three-Dimensional Image, Integral Imaging, Augmented reality

Major Advisor

Bahram Javidi

Associate Advisor

Rajeev Bansal

Associate Advisor

Helena Silva

Field of Study

Electrical Engineering

Degree

Doctor of Philosophy

Open Access

Open Access

Abstract

In recent decades, there have been significant technological advancements in sensors, devices, materials, algorithms, and computational hardware, resulting in extensive improvements for visualization capabilities applied to real world objects. Among these improvements, three-dimensional (3D) imaging technologies have received interest from many research groups and may offer advantages over conventional two-dimensional (2D) approaches. In comparison with 2D sensing techniques, which record only the intensity of the scene, passive 3D imaging also includes depth and directional information. Many techniques for 3D imaging have been proposed, such as holography and interferometry, time-of-flight, two-view based stereoscopy, and multi-view techniques for autostereoscopic 3D imaging, to cite a few.

This dissertation will focus on novel aspects of integral imaging based multi-view 3D imaging systems, 3D information processing and visualization in three separate parts. In the first part, two concepts for integral imaging based dynamic 3D imaging are presented. First, an extended depth-of-focus 3D micro display is presented by using a bifocal liquid crystal lens. Second, a head tracking 3D display is presented by means of proper application of the smart pseudoscopic-to-orthoscopic conversion (SPOC) algorithm.

In the second part, novel 3D imaging systems and 3D image processing approaches are proposed. First, the recent progress of integral imaging based Multidimensional Optical Sensing and Imaging Systems (MOSIS) is presented for object recognition, material inspection, and integrated visualization, etc. Second, 3D profilometric reconstruction using flexible sensing integral imaging with automatic occlusion removal is presented. Third, spatial-temporal human gesture recognition under degraded conditions using 3D integral imaging is presented.

In the third part, approaches for 3D sensing and visualization with Augmented Reality (AR) are presented. First, an AR based approach for optical visualization and object recognition using 3D axially distributed sensing (ADS) is presented. Second, we present an eye fatigue-free 3D augmented display using lenslet based integral imaging. Lastly, a dynamic 3D imaging system based optical see-through augmented reality display with enhanced depth range of a 3D augmented image is presented to reduce the accommodation-convergence mismatch problem.

Download
COinS