Tee, Cheng Siew (2008) Feature selection for content-based image retrieval using statistical discriminant analysis. Masters thesis, Universiti Teknologi Malaysia, Faculty of Computer Science and Information System.
Restricted to Repository staff only
PDF (Table Of Content)
PDF (1st Chapter)
As we known, the very large repository of digital media arise the challenge of various digital search applications. In order to make use of this huge amount of data, effective tools are required for retrieve multimedia information. An image retrieval system is one of the tools that can be used for searching and retrieving images from a large database of digital images. However, there are several challenges and problems need to be considered when applied image retrieval system such as the gap between high-level semantic concept and low-level visual features. This refers to problem of feature selection, which is critical to really solve the gap problem in CBIR. Recently, the most feasible feature selection method is discriminant analysis. Therefore, in this project, we proposed title feature selection in content-based image retrieval using statistical discriminant analysis. In the project, we intended to enhance performance by improve the feature selection process. Besides, we used fuzzy theory in content-based image retrieval to solve the problem of perspective subjectivity of human in image retrieval. The system would be more depends to the human-like and how to response with relevant images that match the concept of current query is always the research question in this project.
|Item Type:||Thesis (Masters)|
|Additional Information:||Supervisor : Assoc. Prof. Dr. Ali bin Selamat|
|Uncontrolled Keywords:||digital media, multimedia information, digital images|
|Subjects:||Q Science > QA Mathematics > QA75 Electronic computers. Computer science|
|Divisions:||Computer Science and Information System (Formerly known)|
|Deposited By:||Ms Zalinda Shuratman|
|Deposited On:||24 Nov 2009 01:50|
|Last Modified:||03 Sep 2012 04:48|
Repository Staff Only: item control page