Home // MMEDIA 2015, The Seventh International Conferences on Advances in Multimedia // View article

Image Retrieval System Based on Combination of Color, Texture and Shape Features

Authors:
Atoany Nazareth Fierro Radilla
Mariko Nakano Miyatake
Karina Perez Daniel
Hector Perez Meana
Francisco García Ugalde
Manuel Cedillo Hernandez

Keywords: CBIR Systems; color descriptor; texture descriptor; shape descriptor; combination methods of descriptors

Abstract:
In the Content-Based Image Retrieval (CBIR) system, an effectiveness of the visual descriptors, such as color, texture and shape, determines a good retrieval performance. Recently, more than two types of visual descriptors are combined to improve the performance of CBIR. The combination method used to combine different types of visual descriptors also plays an important role to obtain a good performance. However, we are aware that researchers have not paid sufficient attention to the combination methods, so in this paper, we focused on combination of schemes of three types of visual features to obtain higher improvement of the retrieval performance. Firstly, several visual descriptors that belong to three types of visual features (color, texture and shaper) are analyzed individually to select a better descriptor from each category. Second, the several combination methods are analyzed to determine a best combination method. The performance of the proposed scheme is compared with some CBIR systems in which more than two different types of descriptors are combined.

Pages: 37 to 43

Copyright: Copyright (c) IARIA, 2015

Publication date: April 19, 2015

Published in: conference

ISSN: 2308-4448

ISBN: 978-1-61208-400-8

Location: Barcelona, Spain

Dates: from April 19, 2015 to April 24, 2015