RT Generic
T1 Image processing and knowledge-based techniques for automated quality improvement of digital images
A1 Casas Muñoz, Pablo
AB With its multidisciplinary nature drawing upon a great variety of areas suchas mathematics, computer graphics, computer vision, visual psychophysics,optics, and computer science, the theory of Image Processing needs to be madeaccessible to practitioners from very diverse backgrounds, from amateurphotographers to specialists in communications, medicine, or biology.The presented thesis focusses on improving quality of images as an automaticprocess, where we regard aesthetic, physical, perceptual and cognitiveapproaches (which respectively result in pleasant, identical, realistic, anddetailed images) as different levels of the same problem: reproduction accuracy.Main contributions include self-contained fundamental material whose value islikely to remain applicable in a rapidly evolving body of knowledge. A basicstrategy followed in its preparation was to provide a seamless integration of wellestablishedtheoretical concepts and their implementation using state-of-the-artsoftware tools.We divide the problem into i) low-level algorithmic routines, which resembleearly-vision stages, where prior knowledge about natural images is implicitly‘coded’ in the algorithm itself; and ii) explicit knowledge representation for highlevelimage processing tasks involving algorithm composition, execution,revision and comparison of candidate solutions.First, we perform a deep theoretical research of classical as well as emergingparadigms in image quality and its two main dimensions, noise and tonereproduction, with emphasis on information theory and signal processing, butalso with inspiration from perceptual sciences and computational photography,thus providing a unifying approach. We focus on edge-preserving smoothingfilters as a simple, yet very powerful low-level image processing tool to deal withnoise reduction and improved tone reproduction by means of extracting intrinsiccomponents of an image. Finally, we develop, prototype and execute selectedlow-level image processing operators in MATLAB.Second, in order to enable end-users to accomplish complex Image Processingtasks while at the same time limiting their cognitive and skill requirements, asystem is provided in which expert’s knowledge is explicitly stated in the formof rules. Developed with classical knowledge-based techniques and finallyimplemented in Java, the proposed system allows easy adaptation to specifictasks by exchanging knowledge bases for different areas like computer vision,remote sensing or medical image analysis.Last, but not least, we are not only interested in solving Image Processingproblems, we also want to capture, understand and share the reasoning behindso that others, including non-expert users, can use and build on it.
YR 2017
FD 2017
LK https://hdl.handle.net/10016/31229
UL https://hdl.handle.net/10016/31229
LA eng
DS e-Archivo
RD 1 sept. 2024