A generic two-dimensional image interpretation system with a high-level user interface
Abstract
Vision systems currently operate in very limited domains compared to the human visual system. We have developed the components of a system for aerial photo analysis which performs region extraction and storage of region features in symbolic data files. The application of rules is then used to analyze the extracted data for object identification. A human interface has been added to this system so that the human's ability to analyze more complex environments can he used to select processing techniques, representation, and control strategies so that the system can perform its tasks in a more general vision domain. Rules are generated automatically from the human input. A new hierarchical partial shape recognition strategy has been developed and incorporated into the system.
Degree
Ph.D.
Advisors
Mitchell, Purdue University.
Subject Area
Electrical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.