Announcement for Downloading full text filePlease respect the Copyright Act.
All digital full text dissertation and theses from this website are authorized the copyright owners. These copyrighted full-text dissertation and theses can be only used for academic, research and non-commercial purposes. Users of this website can search, read, and print for personal usage. In respect of the Copyright Act of the Republic of China, please do not reproduce, distribute, change, or edit the content of these dissertations and theses without any permission. Please do not create any work based upon a pre-existing work by reproduction, Adaptation, Distribution or other means.
URN etd-0906111-165933 Statistics This thesis had been viewed 1924 times. Download 1626 times. Author Sung-En Ku Author's Email Address No Public. Department Communication Engineering Year 2010 Semester 2 Degree Master Type of Document Master's Thesis Language zh-TW.Big5 Chinese Page Count 76 Title ASSESSMENT OF THE FIDELITY METRICS FOR COLOR IMAGES BASED ON THE CONCEPT OF DISTORTION MAP Keyword False Al Color metric Distortion map Hit Rate Hit Rate Distortion map Color metric False Al Abstract High-quality color image reproduction and duplication for many digital image processing systems is very important. In the this research, the author proposes a series of color metrics for assessment, and they are RMS, CIELAB △94, S-CIELAB △94 and PSPNR, trying to find which color metric can accurately predict the response of the human visual system (HVS) to the distortion occurring in the reproduced color image. In this research, color images fidelity metric are evaluated and designed to take only the part of the distortion that is perceivable by the HVS, or the part of the distortion that exceeds just noticeable color difference (JNCD). The validity of the quality metric is verified by using the image distortion maps that are to be obtained empirically from the pairs of the original and reproduced images in the spatio-temporal domain by a number of subjects under a specified viewing condition.
In order to obtain each pixel's distortion value of the reproduction image, the information provided through distortion map is brought into the formula for each metric. All data is conditional screening through the statistical process to get two sets of probability values, hit rate (HR) and false alarm rate (FAR), and to draw the ROC curve. The four metrics demonstrate four different curves, among which PSPNR is found most consistent with human visual system (HVS), which helps the applications of color image processing in different fields in the future.
Advisor Committee Chun-Hsien Chou - advisor
Ray-chin Wu - co-chair
Tan-Hsu Tan - co-chair
Files Date of Defense 2011-07-26 Date of Submission 2011-09-06