Full metadata record
|dc.contributor||Department of Electronic and Information Engineering||en_US|
|dc.publisher||Hong Kong Polytechnic University||-|
|dc.rights||All rights reserved||en_US|
|dc.title||Management of image database with self-organizing map||en_US|
|dcterms.abstract||This dissertation presents management of image database with Self-Organizing Map. Two main contributions are reported: (1) A hybrid system combines image pre-classification and Self-Organizing Map modules for image classification management; (2) A new mixed training method based on Self-Organizing Map for map optimization under non-stationary environment. The hybrid system is designed to firstly classify the images in the database into attentive and non-attentive two categories, and then adaptively extract targeted features from the images with specific properties. Finally, Self-Organizing Map (SOM) is used to cluster the images' features within each category individually so as to implement the image classification management based on content. The conventional Self-Organizing Map can only be trained under stationary environment, in which the dataset has to be fixed if the training process begins, while the map structure cannot be changed once the training process ends. However, in practice trie database may not be permanently invariant for applications, for example, when the training dataset is enlarged, and the map often needs to be adaptively changed. The conventional SOM system cannot simply alter the nodes distribution in the output layer of the map unless destroying the old prototype and training a new one; while if the training dataset contains lots of samples, the whole process will be very time consuming. In this thesis, we propose a new mixed training method that provides a fast algorithm for optimizing the nodes distribution in SOM according to the new coming data.||en_US|
|dcterms.extent||x, 84 leaves : ill. ; 30 cm.||en_US|
|dcterms.isPartOf||PolyU Electronic Theses||en_US|
|dcterms.LCSH||Hong Kong Polytechnic University -- Dissertations.||en_US|
|dcterms.LCSH||Neural networks (Computer science)||en_US|
|dcterms.LCSH||Image processing -- Digital techniques.||en_US|
Files in This Item:
|b23030641.pdf||For All Users (off-campus access for PolyU Staff & Students only)||10.26 MB||Adobe PDF||View/Open|
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item: