Publication Date

8-2017

Date of Final Oral Examination (Defense)

6-30-2017

Type of Culminating Activity

Thesis

Degree Title

Master of Science in Computer Science

Department

Computer Science

Major Advisor

Steven Cutchin, Ph.D.

Advisor

Jerry Alan Fails, Ph.D.

Advisor

Amit Jain, Ph.D.

Advisor

Maria Soledad Pera, Ph.D.

Abstract

Head mounted displays are characterized by relatively low resolution and low dynamic range. These limitations significantly reduce the visual quality of photo-realistic captures on such displays. This thesis presents an interactive view optimized tone mapping technique for viewing large sized high dynamic range panoramas up to 16384 by 8192 on head mounted displays. This technique generates a separate file storing pre-computed view-adjusted mapping function parameters. We define this technique as ToneTexture. The use of a view adjusted tone mapping allows for expansion of the perceived color space available to the end user. This yields an improved visual appearance of both high dynamic range panoramas and low dynamic range panoramas on such displays. Moreover, by providing proper interface to manipulate on ToneTexture, users are allowed to adjust the mapping function as to changing color emphasis. The authors present comparisons of the results produced by ToneTexture technique against widely-used Reinhard tone mapping operator and Filmic tone mapping operator both objectively via a mathematical quality assessment metrics and subjectively through user study. Demonstration systems are available for desktop and head mounted displays such as Oculus Rift and GearVR.

DOI

https://doi.org/10.18122/B2F401

Available for download on Wednesday, August 21, 2019

Files over 30MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS