Cloud Computing and Big Data: Technologies, Applications and Security by Unknown

Author:Unknown
Language: eng
Format: epub
ISBN: 9783319977195
Publisher: Springer International Publishing

2 Related Work

The main objectives of our method are the reduction of features dimensionality, adapted (binary) representation of features and images indexation. In this context, we can categorize three kinds of related works: dimensionality reduction methods, storage structures and general content based retrieval system.

2.1 Content Based Retrieval System

A content based image retrieval system is a computer system for browsing, searching and retrieving images from a large data base of digital images. Several works have been proposed in this area for the domains of commerce, government, academia, and hospitals, where large collections of digital images are being created. Many of these collections are the product of digitizing existing collections of analogue photographs, diagrams, drawings, paintings, and prints. Usually, the only way of searching these collections is done by keyword indexing, or simply by browsing. Morever digital images databases, open the way to content-based searching [11].

In this context, Hirata and Kato [12] proposed an image retrieval system (CBIR) which facilitates the access to images by visual example. In their system, called “query by visual example”, edge extraction is performed on user queries. These edges are matched against those of the database images in a fairly complex process that allows for corresponding edges to be shifted or deformed with respect to each other [13]. In this work, authors did not provide any content-based image indexing mechanism.

Otherwise, The QBIC system [14] presents one of the most notable technics, developed by IBM, for querying by image content. The latter allows a user to compose a query based on a variety of different visual properties such as color, shape, texture, which are semi-automatically extracted from images. This method used partially the R*-tree as an image indexing method [13, 15].

Authors in [12], proposed an image match criteria and a system that uses spatial information and visual features represented by dominant wavelet coefficients. Although their system provides an improved matching over image distance norms, it does not support any index structure. In fact, they mainly focused on efficient feature extraction by using wavelet transform rather than an index structure to support speedy retrieval [1].

The VisualSEEk [16] is a content-based image query system using color regions and spatial layout. In this work, authors proposed an image similarity function, which contains both color feature and spatial components. For image similarity matching, they use intrinsic parts such as colors, region size, spatial location and derived parts such as relative spatial locations. However, the evaluations for derived parameter present very complex operations. The authors used spatial Quad-tree or R-tree for single region query and 2D-string to represent the spatial relationship in an image for multiple region queries. These index structures were devised respectively for each query and were not integrated into a content-based indexing [15].

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.