The Semantic Web project has introduced new techniques for managing information. Data can now be organised more efficiently and in such a way that computers can take advantage of the relationships that characterise the given input to present more relevant output. Semantic Web based search engines can quickly educe exactly what is needed to be found and retrieve it while avoiding information overload. Up until now, search engines have interacted with their users by asking them to look for words and phrases. We propose the creation of a new generation Semantic Web search engine that will offer a visual interface for queries and results. To create such an engine, information input must be viewed not merely as keywords, but as specific concepts and objects which are all part of the same universal system. To make the manipulation of the interconnected visual objects simpler and more natural, 3D graphics are utilised, based on the X3D Web standard, allowing users to semantically synthesise their queries faster and in a more logical way, both for them and the computer.

The effects of immersion with respect to information visualization have rarely been explored. In this thesis, we describe a methodology, two information visualization applications that were developed using the CAVE, and three user studies in order to explore, examine and attempt to quantify the effects of immersion. We focus on three major components of immersion: field of regard (FOR), head-based rendering (HBR), and stereoscopic viewing. We hypothesize that a high degree of FOR will result in increased task performance and user satisfaction when visualizing data represented by scatter and surface plots. We also hypothesize that HBR and stereoscopic viewing will result in increased task performance, but the effects of these components would be greater in the scatter plots than surface plots. We have conducted three user studies with the information visualization applications developed for this research. In the first study, an exploratory pilot study, we observed a trend in favor of using high FOR and HBR. In the second exploratory pilot study, we observed a slight trend in favor of high FOR. In the third study, thirty-two subjects performed tasks using both the scatter plots and surface plots with eight test conditions. We observed trends in favor of high levels of FOR, HBR and stereoscopic viewing in scatter plots, a slight trend in favor of HBR for surface plots, and a significant interaction effect between FOR and HBR in scatter plots for a particular task. / Master of Science

Much research effort has been devoted to producing algorithms that contribute directly or indirectly to the extraction of 3D information from a wide variety of types of scenes and conditions of image capture. The research work presented in this thesis is aimed at three distinct applications in this area: interactively extracting 3D points from a pair of uncalibrated images in a flexible way; finding corresponding points automatically in high resolution images, particularly those of archaeological scenes captured from a freely moving light aircraft; and improving a correlation approach to dense disparity mapping leading to 3D surface reconstructions. The fundamental concepts required to describe the principles of stereo vision, the camera models, and the epipolar geometry described by the fundamental matrix are introduced, followed by a detailed literature review of existing methods. An interactive system for viewing a scene via a monochrome or colour anaglyph is presented which allows the user to choose the level of compromise between amount of colour and ghosting perceived by controlling colour saturation, and to choose the depth plane of interest. An improved method of extracting 3D coordinates from disparity values when there is significant error is presented. Interactive methods, while very flexible, require significant effort from the user finding and fusing corresponding points and the thesis continues by presenting several variants of existing scale invariant feature transform methods to automatically find correspondences in uncalibrated high resolution aerial images with improved speed and memory requirements. In addition, a contribution to estimating lens distortion correction by a Levenberg Marquard based method is presented; generating data strings for straight lines which are essential input for estimating lens distortion correction. The remainder of the thesis presents correlation based methods for generating dense disparity maps based on single and multiple image rectifications using sets of automatically found correspondences and demonstrates improvements obtained using the latter method. Some example views of point clouds for 3D surfaces produced from pairs of uncalibrated images using the methods presented in the thesis are included.

Much research effort has been devoted to producing algorithms that contribute directly or indirectly to the extraction of 3D information from a wide variety of types of scenes and conditions of image capture. The research work presented in this thesis is aimed at three distinct applications in this area: interactively extracting 3D points from a pair of uncalibrated images in a flexible way; finding corresponding points automatically in high resolution images, particularly those of archaeological scenes captured from a freely moving light aircraft; and improving a correlation approach to dense disparity mapping leading to 3D surface reconstructions. The fundamental concepts required to describe the principles of stereo vision, the camera models, and the epipolar geometry described by the fundamental matrix are introduced, followed by a detailed literature review of existing methods. An interactive system for viewing a scene via a monochrome or colour anaglyph is presented which allows the user to choose the level of compromise between amount of colour and ghosting perceived by controlling colour saturation, and to choose the depth plane of interest. An improved method of extracting 3D coordinates from disparity values when there is significant error is presented. Interactive methods, while very flexible, require significant effort from the user finding and fusing corresponding points and the thesis continues by presenting several variants of existing scale invariant feature transform methods to automatically find correspondences in uncalibrated high resolution aerial images with improved speed and memory requirements. In addition, a contribution to estimating lens distortion correction by a Levenberg Marquard based method is presented; generating data strings for straight lines which are essential input for estimating lens distortion correction. The remainder of the thesis presents correlation based methods for generating dense disparity maps based on single and multiple image rectifications using sets of automatically found correspondences and demonstrates improvements obtained using the latter method. Some example views of point clouds for 3D surfaces produced from pairs of uncalibrated images using the methods presented in the thesis are included. / Al-Baath University / The appendices files and images are not available online.