StatusThe thesis was presented on the 14 February, 2018
Approved by NCAA on the 11 May, 2018
Abstract– 1.68 Mb / in romanian
– 1.78 Mb / in english
ThesisCZU 004.382.76: 004.932:77(043)
6.80 Mb /
Thesis structure: The thesis contains Introduction, 3 chapters, general conclusions and recommendations, bibliography of 101 titles. The main text amounts to 107 pages, includes 66 figures, 30 tables, 37 formulas, and 5 appendixes. The obtained results of the thesis were published in 7 scientific papers.
Aim of research is to develop a new no or zero (NR) reference subjective quality assessment model and framework that would enable the smartphones industry prediction of the PIQ.
The objectives of thesis include the design of NR subjective and objective image quality metrics based on extensive visual tests experiments and evaluation by SW tool VIQET that measure the perceived image quality of smartphones users.
Scientific novelty and originalityof the obtained results is reflected in a new approach which predicts the PIQ by extraction of the classic, objective IQAs (brightness, contrast, color saturation and sharpness) to new subjective IQ criteria that evaluated by VIQET. This is a new assessment process based on a new diploid model, which aims to predict how the smartphones users perceived image quality.
Important scientific solved problem consists in elaborating a new diploid model for PIQ assessment that describes how to extract the standard IQAs which used in large screens TVs into new IQAs of SW tool VIQET. This can be used by smartphone producers and vendors in order to shorten „time to market”, or by any image quality experts, including in the academia in order to reduce time and cost of PIQ assessment process of smartphones with small HD displays in comparison to the process that based on many human physical tests.
Theoretical significance is supported by analyzing, specifying and defining the theoretical principles and new PIQ criteria which implemented in the SW tool VIQET in order to predict the expected MOS in HVTs. The new model determines the relationship between the standard IQAs values and the VIQET PIQ criteria.
Applicative value of the work is determined by the developed model, which has huge potential for smartphone industry and users in reducing significantly the time and cost of PIQ assessment.
Implementation of results: the obtained results are used in „Ort Braude College of Engineering” in Israel, and can further be used by students and researchers in image processing. Also, an Israeli company (ASI – Applied Spectral Imaging Ltd.) evaluated the PIQ framework and found the benefits of PIQ framework for IQ improvement of images taken by medical spectral imaging system.