Date of Graduation

2007

Document Type

Dissertation/Thesis

Abstract

This dissertation examines the coupling of GIS and immersive visualization (IV). This study was premised on the hypothesis that linking IV and GIS could potentially enhance a user’s visual-cognitive capacity to perceive, analyze, and understand complex geospatial data. A loosely-coupled GIS-IV system was developed and subsequently tested by professional soil scientists in soil boundary mapping and soil map update case studies. Soil boundary mapping is essentially a visualization enterprise involving the creation of cognitive models representing the relationship between soil and observable environmental features. A GIS-IV conceptual model was formulated, and an operational system was designed and implemented. The GIS-IV model seeks to bring the cognitive and logical semantic worlds closer together, and move GIS toward experiential immersion within the geospatial and mapped data. The model places the GIS-IV system within a framework that emphasizes significant user-computer interactivity, multidimensional representation, and experiential knowledge creation and understanding. The GIS-IV framework supports geocomputational analysis, and facilitates ‘visual and spatial thinking’. In addition, the GIS-IV model bridges the historical divide between GIS and advanced geovisualization methods, and extends the visualization capability of GIS well beyond traditional cartographic 2D mapping. The GIS-IV system was implemented using commercial-off-the-shelf (COTS) software, a stereoscopically-enabled multi-user immersive Cave Automatic Virtual Environment (CAVE), a pen-based Tablet PC, and an enterprise geodatabase server. The GIS-IV system was supported by a robust geospatial data management system, geospatial analysis, and geovisual analytical capabilities. In addition, the system is scalable, extensible, and flexible, and facilitates and encourages geocollaboration between researchers. A user-based and task-based use and usability testing of the GIS-IV system in two soil mapping applications involved several collaborating soil scientists, and revealed very positive reactions, considerable commonality in viewpoint, and occasional varying user perceptions of the system and experience of performing collaborative ‘virtual’ soil mapping. Overall, the participants’ questionnaire responses reflected positively on the use of the GIS-IV system for virtual soil boundary mapping and soil map revision. In particular, the ability of the system to support ‘same place-same time’ geocollaborative interpretative image analysis, a ‘go anywhere’ capability, and an immersive and experiential interpretation and mapping environment that provided access to physically inaccessible or trespass prohibited areas as well as multiple aerial imagery and geospatial datasets were identified as the main strengths of the system. The participants found the GIS-IV system to be more intuitive than traditional soil mapping practice, and capable of improving the speed and quality of soil mapping. The quality and accuracy of the virtual soil map products were examined using comparative visual analysis, a confusion matrix, a fuzzy agreement matrix, and through geospatial ‘error’ modeling methods. The fuzzy-based approach is sensitive to the imprecision in human reasoning used in soil mapping and is recommended as the best approach for assessing the quality of the soil map products.

Share

COinS