CBIR .. what does this stand for ?!
CBIR stands for Content Based Image Retrieval.
and in brief CBIR is a technique used in image retrieval from large image databases, instead of relying on on metadata such as captions or keywords in retrieving images which is expensive while saving images and inaccurate in retrieval.
"Content" may be ambiguous, to be clear content of an image may be colors in it, textures, shapes or any other information that can be extracted from an image.
The following figure shows a general structure of CBIR systems
CBIR systems always consists of 2 phases the offline phase and the offline phase
where “offline” indexing phase is displayed in the bottom part. Usually the visual contents of the images in the database are extracted and described by multidimensional feature vectors. A feature database is formed using the feature vectors of the images in the database. The “online” content-based retrieval is displayed in the upper part. Users provide the retrieval system with query example images, which are used to retrieve images.
The system then extracts the feature vectors from the example images. The similarities/distances between the feature vectors of the query example and those of the images in the database are calculated. The online and offline phase interact with a collection of multimedia items (images, videos, etc.) from a multimedia database. The query provided by the user can be an example image, region, sketch, humming, or text .
Medicine is one of the most potential application area.
With medical imaging techniques such as X-Ray, computer tomography, magnetic resonance imaging, and ultrasound, the amount of digital images that are produced in hospitals is increasing incredibly fast. Thus the need for systems that can provide efficient retrieval of images of particular interest is becoming very high.
Since for very large database personally describing and annotating every image with text indices is time-consuming and impractical (TBIR - Text Based Image Retrieval or also called Metadata approach), CBIR seems to be a good approach.
Researchers are still working in this field, it's a new born field.
and thus we'll participate in researches and applications in this field isA :)
Resources:
1- Ivica Dimitrovski, Dejan Gorgevik, Suzana Loskovska, WEB-BASED MEDICAL IMAGE RETRIEVAL SYSTEM
2- http://en.wikipedia.org/wiki/Content-based_image_retrieval
CBIR stands for Content Based Image Retrieval.
and in brief CBIR is a technique used in image retrieval from large image databases, instead of relying on on metadata such as captions or keywords in retrieving images which is expensive while saving images and inaccurate in retrieval.
"Content" may be ambiguous, to be clear content of an image may be colors in it, textures, shapes or any other information that can be extracted from an image.
The following figure shows a general structure of CBIR systems
CBIR systems always consists of 2 phases the offline phase and the offline phase
where “offline” indexing phase is displayed in the bottom part. Usually the visual contents of the images in the database are extracted and described by multidimensional feature vectors. A feature database is formed using the feature vectors of the images in the database. The “online” content-based retrieval is displayed in the upper part. Users provide the retrieval system with query example images, which are used to retrieve images.
The system then extracts the feature vectors from the example images. The similarities/distances between the feature vectors of the query example and those of the images in the database are calculated. The online and offline phase interact with a collection of multimedia items (images, videos, etc.) from a multimedia database. The query provided by the user can be an example image, region, sketch, humming, or text .
Medicine is one of the most potential application area.
With medical imaging techniques such as X-Ray, computer tomography, magnetic resonance imaging, and ultrasound, the amount of digital images that are produced in hospitals is increasing incredibly fast. Thus the need for systems that can provide efficient retrieval of images of particular interest is becoming very high.
Since for very large database personally describing and annotating every image with text indices is time-consuming and impractical (TBIR - Text Based Image Retrieval or also called Metadata approach), CBIR seems to be a good approach.
Researchers are still working in this field, it's a new born field.
and thus we'll participate in researches and applications in this field isA :)
Resources:
1- Ivica Dimitrovski, Dejan Gorgevik, Suzana Loskovska, WEB-BASED MEDICAL IMAGE RETRIEVAL SYSTEM
2- http://en.wikipedia.org/wiki/Content-based_image_retrieval
