Please use this identifier to cite or link to this item: http://hdl.handle.net/10603/519783
Title: Effective feature extraction and image indexing for web image mining
Researcher: Karthikeyan R
Guide(s): Celine Kavida A
Keywords: Computer Science
Computer Science Information Systems
Engineering and Technology
Image indexing
Memory Optimization
Web image mining
University: Anna University
Completed Date: 2021
Abstract: A good web image mining system is expected to provide users, the underlying knowledge and patterns in images which include image storage, image processing, feature extraction, image indexing and retrieval, pattern and knowledge discovery. So, Image mining systems needs an efficient mechanism for image retrieval conventional databases using Primary and Secondary indexes used currently the retrieval system for similarity hidden phase. In this case indexing has been done in similarity space. Indexing techniques ranging from standard methods such as inverted access methods and multidimensional methods. A specific indexing scheme is necessary for web image mining which presents and efficient color indexing scheme for similarity based retrieval system with the advancement in new image retrieval mechanism, gives better performance in retrieval techniques towards accuracy, time and memory utilization. Good indexing results in better classification and clustering of large image databases. Hence, the indexing mechanism is very important step in web image mining. Accessing newlinelarge image databases which are available in the web portal needs opted Indexing structure instead of reducing the contents of different kind of database to process quickly. It paves a path towards the increase in the number of efficient image retrieval techniques and numerous researches in the field of image indexing in large image data sets. Normally the image retrieval is facing difficulties like a) Merging the diverse representations of Images and its Indexing is a challenging one b) The low level visual characters and semantic characters associated with an image are indirect proportional one and c) Noisy and less accuracy in extraction of image information (semantic and predicted attributes). This research work clearly focuses and takes the base of reverse engineering and de-normalizing concept by looking back on, how the data can be stored effectively. So the retrieval becomes easy and too fast. The following works have been carried out and results
Pagination: xvii , 125 p.
URI: http://hdl.handle.net/10603/519783
Appears in Departments:Faculty of Information and Communication Engineering

Files in This Item:
File Description SizeFormat 
01_title.pdfAttached File446.25 kBAdobe PDFView/Open
02_prelim_pages.pdf2.49 MBAdobe PDFView/Open
03_content.pdf772.02 kBAdobe PDFView/Open
04_abstract.pdf643.53 kBAdobe PDFView/Open
05_chapter 1.pdf1.51 MBAdobe PDFView/Open
06_chapter 2.pdf804.22 kBAdobe PDFView/Open
07_chapter 3.pdf1.34 MBAdobe PDFView/Open
08_chapter 4.pdf1.79 MBAdobe PDFView/Open
09_chapter 5.pdf1.5 MBAdobe PDFView/Open
10_annexures.pdf169.01 kBAdobe PDFView/Open
80_recommendation.pdf126 kBAdobe PDFView/Open
Show full item record


Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).

Altmetric Badge: