Mobile media search engine

This Final Year Project report describes the analysis, theoretical review and implementation of an image near duplicate detection search engine. It also explores the feasibility of improving image annotation result to enable more precise media search engine using time information from image...

Full description

Saved in:
Bibliographic Details
Main Author: Cheah, Aaron Wei Loong.
Other Authors: Yap Kim Hui
Format: Final Year Project
Language:English
Published: 2011
Subjects:
Online Access:http://hdl.handle.net/10356/45224
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:This Final Year Project report describes the analysis, theoretical review and implementation of an image near duplicate detection search engine. It also explores the feasibility of improving image annotation result to enable more precise media search engine using time information from images. Image near duplicate detection has been an increasingly interesting research field due to rapid growth in personal image collection of mobile device users. It has a high potential in helping users to annotate image by retrieving near duplicate images as well as retrieving similar images to build albums. It also allows users to search for visually similar images using image-based search. The near duplicate image detection part of project was based on extracting features using Scale Invariant Feature Transform (SIFT) and feature matching method to retrieve near duplicate images. The proposed method has shown commendable result on the system with high precision versus recall based on our image database compared with other methods. A more detailed review on the database used, chosen method and other available methods is discussed.