Rochester Institute of Technology
Reasoning with visual knowledge in an object recognition system
Abstract
dc:description.abstractThe impact of artificial intelligence on computer vision has provided various perspectives and approaches to solving problems of the human visual system. Some of the symbolic processing and knowledge-based techniques implemented in vision systems represent a meaningful extension to the low-level, algorithmic processing which has been emphasized since the advent of the computer vision field. The higher-level processes attempt to capture the essence of visual cognition, specifically by encompassing a model of the visual world and the reasoning processes that manipulate this stored visual knowledge and environmental cues. This thesis includes a discussion of existing computer vision systems surveyed from a high-level perspective. The goal of this thesis is to develop a high-level inference system that implements reasoning processes and utilizes a visual memory model to achieve object recognition in a specific domain. The focus is on symbolically representing and reasoning with high-level knowledge using a frame-based approach. The organization and structuring of domain knowledge, reasoning processes and control and search strategies are emphasized. The implementation utilizes a frame package written in Prolog.
Degree
thesis:*- Level thesis:degree_level
- Thesis
- Discipline thesis:degree_discipline
- Computer Science (GCCIS)
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Wojnowski, Christine
- Contributors dc:contributor
-
- Not Listed
Subjects
dc:subject × 3Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.rit.edu/theses/586