15-06-2013, 11:39 AM
ACTIVE LEARNING METHODS FOR INTERACTIVE IMAGE RETRIEVAL
ACTIVE LEARNING.docx (Size: 44.3 KB / Downloads: 18)
Project Scope:
This project is mainly designed considering the
scenarios for comparing the images dynamically. Active
learning methods are used to interact with the user.
Although it is used for image comparison which fulfills
exact requirements according to the user. But it is not
used in case of video applications. Finally, we
compare our method to existing ones using real
scenario on large databases.
Design and Implementation Constraints:
_ The software is designed in such a way that the user can easily
interact with the screen because they are GUI. The customer
has to enter the image dynamically and select any folder then
start comparing the images one by one. So the set of similar
images are displayed.
User Documentation:
_ In our user manual we are going perform image comparison
in a very interactive way, which can be understandable by a
new person who is going to use it. If a new person is using it,
necessary information will be provided. We are going to
explain each and every step clearly about our product so that
any user can easily understand it.
SYSTEM ANALYSIS AND DESCRIPTION
Existing System:
In the existing system the CBIR method faced a lot of
disadvantage in case of the image retrieval. The
following are the main disadvantage faced in case of
the medical field - Medical image description is an
important problem in content-based medical image
retrieval. Hierarchical medical image semantic
features description model is proposed according to
the main sources to get semantic features currently.
Hence we propose the new algorithm to over come
the existing system. In existing system, Images were
first annotated with text and then searched using a
text-based approach from traditional database
management systems
Proposed System:
In the proposed system we use the retin technique. In this
technique the user gives his input in the form of an image.
We then check for the images in the training set rather than
going to the database to search the images. The training set
contains the images that are frequently searched by the user.
Then all the relevant images are compared and the top rank
similar images are displayed as the output. Here the
efficiency and accuracy are increased and the drawbacks of
the existing system are overcome.
Image Retrieval:
An image retrieval system is a computer system for browsing,
searching and retrieving images from a large database of digital
images. Most traditional and common methods of image retrieval
utilize some method of adding metadata such as captioning,
keywords, or descriptions to the images so that retrieval can be
performed over the annotation words. Manual image annotation is
time-consuming, laborious and expensive; to address this, there
has been a large amount of research done on automatic image
annotation. Additionally, the increase in social web applications
and the semantic web have inspired the development of several
web-based image annotation tools.
Other Query Methods
Other query methods include browsing for example images,
navigating customized/hierarchical categories, querying by
image region (rather than the entire image), querying by
multiple example images, querying by visual sketch, querying
by direct specification of image features, and multimodal
queries (e.g. combining touch, voice, etc.)
CBIR systems can also make use of relevance feedback, where
the user progressively refines the search results by marking
images in the results as "relevant", "not relevant", or
"neutral" to the search query, then repeating the search with
the new information.