A Semantic Approach to Text-based Image Retrieval Using a Lexical Ontology

• Main Author: Engku Fadzli Hasan, Syed Abdullah
• Format: Conference or Workshop Item
• Language: English; English
• Published: 2010
• Subjects: T Technology (General); TA Engineering (General). Civil engineering (General)
• Online Access: http://eprints.unisza.edu.my/116/1/FH03-FIK-14-00928.pdf; http://eprints.unisza.edu.my/116/2/FH03-FIK-14-02044.png; http://eprints.unisza.edu.my/116/

This research proposed a novel semantic approach to text-based image retrieval based on a lexical ontology called OntoRo. The approach aims to narrow down the semantic gap between visual content and the richness of human semantics by using keywords-based semantic image indexing. The approach proposed involves: (a) semantic image indexing; (b) semantic search; and (c) semantic image visualization. This paper focuses on the semantic image indexing method and implements semantic search method as the evaluation process. A new concept of Semantic DNA is introduced as an indexing method which will help to retrieve semantic results. The Semantic DNA is extracted from the existing ontological structure of OntoRo. It is the key element in this approach which will be used throughout the whole research.