Skip to Main Content
Frequently Asked Questions
Submit an ETD
Global Search Box
Need Help?
Keyword Search
Participating Institutions
Advanced Search
School Logo
Files
File List
12795.pdf (3.83 MB)
ETD Abstract Container
Abstract Header
Information Theoretic Methodology For Retrieving Similar GIS Rasters
Author Info
Venkatakrishnan, Subashini
Permalink:
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1416233818
Abstract Details
Year and Degree
2014, MS, University of Cincinnati, Engineering and Applied Science: Computer Science.
Abstract
A few decades ago most of the stored digital information was for text and numeric data but recently the volume of data representing images or similar structures such as maps and rasters has been continuously increasing. Such image databases are becoming increasingly useful in medical, geographical, scientific and engineering domains. A very large number of such digital images are getting generated every day and are being saved for analysis and retrieval. There is a need for retrieval of relevant images to conduct studies and research for development in every domain. There are mainly two types of methods for querying image collections and retrieving images that are similar to a query image. First method retrieves images based on the keywords or text tags associated with images. Due to its strong limitations, content based image retrieval methods have been developed and these retrieve images based on the information contained within the images. Features of an image such as color, shape, texture etc. form the basis of content based image retrieval. Our image collection consists of rasters of a terrain and each pixel characterizes the terrain characteristic of its location. The goal of our methodology is to retrieve those rasters that contain terrain characteristics similar to those in the query image. The retrieval methodology should consider these features irrespective of their location or orientation in a terrain raster. As these images contain features similar to texture features, we have devised a methodology to perform content based image retrieval using entropy-based features existing in various parts of the terrain in each image. The goal is to capture terrain characteristics in a query image and then match them with target images independent of location and orientation in the matched images. It is shown in this thesis that our methodology produces very good results. The only validation method available to us is the human subjective evaluation of the retrieved images and enough examples are shown to illustrate the results of our methodology. Our algorithm is compared with the previous work done in this area and the similarities and differences are mentioned. Experiments are conducted using different numbers and types of features extracted from the images and the results are presented in the thesis.
Committee
Raj Bhatnagar, Ph.D. (Committee Chair)
Yizong Cheng, Ph.D. (Committee Member)
Tomasz Stepinski, Ph.D. (Committee Member)
Pages
101 p.
Subject Headings
Computer Science
Keywords
CBIR
;
texture
;
GIS
;
similar images
;
rasters
Recommended Citations
Refworks
EndNote
RIS
Mendeley
Citations
Venkatakrishnan, S. (2014).
Information Theoretic Methodology For Retrieving Similar GIS Rasters
[Master's thesis, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1416233818
APA Style (7th edition)
Venkatakrishnan, Subashini.
Information Theoretic Methodology For Retrieving Similar GIS Rasters.
2014. University of Cincinnati, Master's thesis.
OhioLINK Electronic Theses and Dissertations Center
, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1416233818.
MLA Style (8th edition)
Venkatakrishnan, Subashini. "Information Theoretic Methodology For Retrieving Similar GIS Rasters." Master's thesis, University of Cincinnati, 2014. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1416233818
Chicago Manual of Style (17th edition)
Abstract Footer
Document number:
ucin1416233818
Download Count:
382
Copyright Info
© 2014, all rights reserved.
This open access ETD is published by University of Cincinnati and OhioLINK.