Date of Award

8-2016

Degree Type

Thesis

Degree Name

Master of Science in Electrical and Computer Engineering (MSECE)

Department

Electrical and Computer Engineering

First Advisor

Amy R. Reibman

Committee Chair

Amy R. Reibman

Committee Member 1

Jianghai Hu

Committee Member 2

Michael D. Zoltowski

Abstract

Digital images are widely used in our daily lives and the quality of images is important to the viewing experience. Low quality images may be blurry or contain noise or compression artifacts. Humans can easily estimate image quality, but it is not practical to use human subjects to measure image quality in real applications. Image Quality Estimators (QE) are algorithms that evaluate image qualities automatically. These QEs compute scores of any input images to represent their qualities. This thesis mainly focuses on evaluating the performance of QEs. Two approaches used in this work are objective software analysis and the subjective database design.

For the first, we create a software consisting of functional modules to test QE performances. These modules can load images from subjective databases or generate distortion images from any input images. Their QE scores are computed and analyzed by the statistical method module so that they can be easily interpreted and reported. Some modules in this software are combined and formed into a published software package: Stress Testing Image Quality Estimators (STIQE).

In addition to the QE analysis software, a new subjective database is designed and implemented using both online and in-lab subjective tests. The database is designed using the pairwise comparison method and the subjective quality scores are computed using the Bradley-Terry model and Maximum Likelihood Estimation (MLE). While four testing phases are designed for this databases, only phase 1 is reported in this work.

Download

Share

COinS