07-07-2012, 01:24 PM
Content Based Image Retrieval
Content Based Image Retrieval.pdf (Size: 536.93 KB / Downloads: 229)
ABSTRACT
This technique is used to retrieve images from large amount of
image collection by using primitive features of images like color, shape, texture etc. It avoids
the annotation of name with each image as in image retrieval using textual metadata. The
main requirements of CBIR systems are techniques for feature extraction, query
specification, user interactions etc.Techniques for feature extraction are described with
currently existing techniques like color retrieval, shape retrieval and texture retrieval. The
importance of user interaction and different methods for it is also described. The performance
evaluation methods and practical applications of CBIR are also discussed in this topic.
Introduction
Advances in data storage and image acquisition technologies have enabled
the creation of large image datasets. In order to deal with these data, it is necessary to develop
appropriate information systems to efficiently manage these collections. Image searching is
one of the most important services that need to be supported by such systems. In general, two
different approaches have been applied to allow searching on image collections: one based on
image textual metadata and another based on image content information.
Architecture of CBIR Systems
Figure shows a typical architecture of a content-based image retrieval
system. Two main functionalities are supported: data insertion and query processing.
The data insertion subsystem is responsible for extracting appropriate features from images
and storing them into the image database (see dashed modules and arrows). This process is
usually performed off-line.
Interfaces to obtain the users’ query demands
As in any retrieval system, the retrieval process begins with a query
requirement. Hence, it is crucial to capture the users’ query demands accurately and easily.
For current CBIR systems, the image retrieval process is usually carried out through an
example image provided by the user, called query-by-example. However, users cannot
always submit an example image to the retrieval system. Typically, current CBIR systems
solve this problem by offering an interface to specify or choose some primitive features for
providing an example image. For instance, when using IBM’s QBIC system.
Techniques to efficiently index and store metadata.
For a large image collection, the storage space required for the metadata is
usually non-trivial. A CBIR system must possess efficient techniques to compress the metadata.
Also, retrieving images and other visual resources, such as video, demands that
standards be set to describe the metadata. When a query is processed against a large image
database, it is often unacceptable to compare the similarity between the query image and all
the images, one by one.