In search of an objective measure for the perceptual quality of printed Images
doctoral thesis
This thesis is concerned with the topic of perceptual image quality. Our visual communication is mainly based on images, either natural or synthetic. Humans are able to judge whether an image is a good image. However, for many applications it would be faster and cheaper to estimate the quality of images using a computer. The major goal of this thesis is to find an objective measure for the quality of a printed image that corresponds to perceptual image quality. In addition to this major goal some secondary goals are defined as: 1) improving insight in the complex topic of perceptual quality, 2) determining the relation between perceptual attributes and image quality and 3) developing methods for color image processing. Two different perceptual attributes are investigated: sharpness and color contrast. Measures are developed that can be used to obtain an objective prediction of sharpening, smoothing and sharpness. In the color experiments the color of the images was changed with a gamma manipulation and a chroma scaling. These manipulations had a significant effect on perceptual quality. In the last part of this thesis it was evaluated how the gamut, that is the enveloppe of colors in color space that determine the limitations of the displaying device, can be incorporated into achromatic enhancement of color images and how this affects image quality. In this way, some of the differences between images displayed on a monitor and printed images can be taken into account.
TNO Identifier
12884
ISBN
90-76702-04-7