Return to search

A rule-based expert system for image segmentation /

A major problem in robotics vision is the segmentation of images of natural scenes in order to understand their content. This thesis presents a new solution to the image segmentation problem that is based on the design of a rule-based expert system. General knowledge about low level properties of an image is formulated into production rules. A number of processes employ the rules to segment the image into uniform regions and connected lines. In addition to the knowledge rules, a set of control rules are also employed. These include meta-rules that embody inferences about the order in which the knowledge rules are matched. They also include focus of attention rules that determine the path of processing within the image. A third set of rules contains the strategy rules which are data-driven inferences about the control rules. They dynamically modify the processing strategy. Different rule ordering and focus of attention strategies are selected according to a set of performance parameters. These measure the quality of the segmentation output at any point in time. Experiments with the knowledge rules resulted in an optimal set based on output quality and processing efficiency. Overall system performance is shown to be qualitatively and quantitatively superior to previous segmentation algorithms.

Identiferoai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:QMM.76723
Date January 1983
CreatorsNazif, Ahmed M.
PublisherMcGill University
Source SetsLibrary and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada
LanguageEnglish
Detected LanguageEnglish
TypeElectronic Thesis or Dissertation
Formatapplication/pdf
CoverageDoctor of Philosophy (Department of Electrical Engineering.)
RightsAll items in eScholarship@McGill are protected by copyright with all rights reserved unless otherwise indicated.
Relationalephsysno: 000167503, proquestno: AAINK64529, Theses scanned by UMI/ProQuest.

Page generated in 0.0018 seconds