𝔖 Bobbio Scriptorium
✦   LIBER   ✦

Intermediate Annotationless Dynamical Object-Index-Based Query in Large Image Archives with Holographic Representation

✍ Scribed by Javed I. Khan


Publisher
Elsevier Science
Year
1996
Tongue
English
Weight
800 KB
Volume
7
Category
Article
ISSN
1047-3203

No coin nor oath required. For personal study only.

✦ Synopsis


This paper presents a new parallel and distributed associative network-based technique for content-based image retrieval denoted by a set of pixels. It is expected that a match (CBIR) with dynamic indices. Unlike any prior artificial assoshould be based on the similarity between these objects.

ciative networks (AAM), this new associative search network

For instance, a radiologist can request organs that are has the unique ability to explicitly focus on any subset of pixels not of the normal shape like the one in a given example in the image. It can also provide a feedback meta-quantity on image. A detective often needs to find if there is any rethe quality of outgoing information. The network is founded corded image which matches some interesting ridge paton a bi-modal representation of information elements which terns of a fingerprint at hand. However, the problem is in addition to basic information also includes meta-states. Its that these image objects and concepts are perceptual computational model has been derived from optical holograand subjective.

phy. These unique capabilities coupled with usual advantages of associative computing (adaptability, efficiency, ability to cope

Nature of Image Information

with imprecision, parallel and distributed mode of computation) now for the first time make it possible to realize a CBIR Image information is representationally sparse and distechnique based on associative computing. This new CBIR tributed and the process of its perceptualization involves strategy provides an inquirer greater flexibility to independently complex abstraction. The more non-graphical 1 an image and dynamically construct object-indices without depending is (such as a medical image), the more the concepts in it, on the fixed, predefined ad hoc indices used by traditional CBIR objects (such as a tumor) and their relationships (such as approaches. The paper presents the mechanism, architecture, anatomical, compositional, or spatial relations), tend to be and performance of an image archival and retrieval system realized with this new network.