Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.485591
Title: Developing Image Quality Metrics of Liquid Crystal Displays: Internal Reference Approach
Author: Kim, Youn Jin
Awarding Body: University of Leeds
Current Institution: University of Leeds
Date of Award: 2007
Availability of Full Text:
Access through EThOS:
Abstract:
Recently, many kinds of mobile displays have become increasingly popular. They are widely used under a variety of ambient illumination levels, including bright outdoor viewing conditions. Manufacturers of mobile displays require new data and computational models which consider the effect of ambient illumination in order to develop higher quality products. The aim of this thesis is to analyse the ability of the human visual system to evaluate the quality of images reproduced on small-size mobile display media, especially for 2-inch liquid crystal display, and to computationally enhance quality of images under a wide range of ambient illumination levels. Specifically, the following four major issues were investigated: analysis of psychophysical attributes affecting image-quality evaluation, quantification of image quality using measurable image-properties, measurement of the ambient-illumination adaptive contrast sensitivity function, and image enhancement which discounts the effect of ambient-illumination. Since small sized mobile display was mainly focused on in the current research, the results can be used for a limited range of display size. Effects 'of display size were remained as future study. Firstly, typical variations of mobile display media that exist in the real world were simulated using a device characterisation technique and psychophysical attributes (e.g. naturalness, clearness, sharpness, contrastness, colourfulness and preference) affecting the image quality evaluation process were analysed. The naturalness and clearness were found to be the most statistically important ( psychophysical-attributes for modelling image quality. As ambient illumination level increases, the significance of clearness (which is attributed to both contrast and sharpness) increased and image quality was impaired exponentially. Secondly, image quality was modelled using the measurable image properties memory colour reproduction ratio (MCRR), mean chroma and 95th percentile luminance - regarding to the physical characteristics of display media. A computational image-colour quality (ICQ) model was developed that consists of three cognitive processes: local and global assessments and scoring the ICQ. The effect of outdoor illumination on the ICQ model estimate could' be 'quantified as an exponential decay function of illuminance (Ix). Thirdly, to estimate the change in image contrast under a variety of ambient illumination levels, the contrast sensitivity function (CSF) of the human visual system was psychophysically measured using the contrast threshold detection method. As the ambient illumination level increased, the overall contrast sensitivity decreased and the spatial. frequency where the maximum contrast sensitivity occurred was shifted towards a lower frequency. Finally, the loss in contrast sensitivity was quantified and compensated for by increasing the amplitude of Fourier transform of a given image. An. ambient- . illumination adaptive weighting function was applied to the' amplitude increase, . according to the normalised CSF difference between the reference (dark) and a given target ambient illumination level. This IEA method resulted in a considerable image· quality enhancement under outdoor viewing conditions, as verified through a set of psychophysical experiments.
Supervisor: Not available Sponsor: Not available
Qualification Name: University of Leeds, 2007 Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.485591  DOI: Not available
Share: