A study on a content-based image retrieval technique for Chinese paintings

Pages172-188
Date05 February 2018
Published date05 February 2018
DOIhttps://doi.org/10.1108/EL-10-2016-0219
AuthorChia-Ching Hung
Subject MatterInformation & knowledge management,Information & communications technology,Internet
A study on a content-based image
retrieval technique for
Chinese paintings
Chia-Ching Hung
Department of Information Management, National Chung Cheng University,
Minhsiung, Taiwan
Abstract
Purpose The purpose of this study is to build a database of digital Chinese painting images and use
the proposed techni que to extract image and texture in formation, and search images simi lar to the query
image based on colour hi stogram and texture fea tures in the database. T hus, retrieving image s by this
image technique is exp ected to make the retrieval of C hinese painting images more pr ecise and convenient
for users.
Design/methodology/approach In this study, a technique is proposed that considers spatial
information of colours in addition to texture feature in image retrieval. This technique can be applied to
retrieval of Chinese paintingimages. A database of 1,200 digital Chinese paintingimages in three categories
was built, including landscape, ower and gure. The authors develop an image-retrieval technique that
considerscolour distribution, spatial informationof colours and texture.
Findings In this study, a databaseof 1,200 digital Chinese painting images in threecategories was built,
including landscape, ower and gure. An image-retrieval technique was developed that considers colour
distribution, spatialinformation of colours and texture. Through adjustment of featurevalues, this technique
is able to processboth landscape and portrait images.This technique also addresses liubai (i.e.blank) and text
problems in the images. The experimental results conrm high precision rate of the proposed retrieval
technique.
Originality/value In this paper,a novel Chinesepainting image-retrieval technique is proposed.Existing
image-retrievaltechniques and the features of Chinese painting are used to retrieveChinese painting images.
The proposed technique can exclude less important image information in Chinese painting images for
instance liubaiand calligraphy while calculating the feature values in them. The experimentalresults conrm
that the proposed technique delivers a retrieval precision rate as highas 92 per cent and does not require a
considerable computing power for feature extraction. This technique can be applied to Web page image
retrievalor to other mobile applications.
Keywords Digital libraries, Digital archives, Image retrieval, Chinese painting images,
Colour histograms
Paper type Research paper
Introduction
With the advancement of information technology and digital image processing techniques,
recent years have seen a proliferation of digital images available on the internet. How to
retrieve images of interest froma huge image database quickly and accurately is, therefore,
very important. General image-retrieval techniques are based on keyword search, which
requires manual input of descriptive information about each image rst. This retrieval
method is in fact very laborious. Therefore, it is necessary to develop a fast, accurate and
labour-savingimage-retrieval system that allows users to retrieveimages of interest from an
image database.
EL
36,1
172
Received13 October 2016
Revised4 April 2017
Accepted12 June 2017
TheElectronic Library
Vol.36 No. 1, 2018
pp. 172-188
© Emerald Publishing Limited
0264-0473
DOI 10.1108/EL-10-2016-0219
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0264-0473.htm
Many museums of art and history have enormous collections of digital images of
historical artefactsavailable for viewing online. This is a part of their effort to build a digital
museum that allows people to view artefacts from across the world at home. Many artists
are also exhibiting and sellingtheir works online. Therefore, how to index, view and retrieve
digital images of art works effectivelyis an important issue that needs to be addressed with
no delay (Jia and Wang, 2004;Jiang et al., 2006;Shen, 2009;Sheng and Jiang, 2014;
Wallraven et al., 2009).
Content-based image retrieval (CBIR) has been extensively studied and discussed.
Various feature-extraction schemes have been proposed, and most schemes are based on
colour (Condorovici et al., 2015;Lu and Chang, 2007), shape (Nam et al., 2008), texture
(Huang and Dai, 2004) or colour histogram(Kim and Chung, 2006;Smeulders et al., 2000).
Chinese painting has its particularities in painting techniques and kinds. It cannot
directly apply the existing image-retrieval technology, so search results are limited. In this
study, the analysis of the characteristics of Chinese painting images will help to nd the
appropriate features to increase search accuracy; considering the future of Big Data,
the number of digital paintings will only increase. In additionto the accuracy of the search,
the reduction is also anotherkey point (Figure 1).
Chinese painting covers a wide variety of forms of painting. These forms of painting
have some common features, including colour and texture (Jiang et al., 2006;Sheng and
Jiang, 2014). A colour histogramis an important feature that can be used for image retrieval.
As colour distribution is not considered in colour histogram-based image retrieval, the
retrieved images may have similar colours but different colour distributions. Precision of
retrieval will be affectedas a result. In this study, annular colour histograms were used. The
method is to draw concentric circles around the centre of the image to record colour
information. Additionally, the method will be modied as proposed by Sun et al. (2006) for
Figure 1.
Digitalizedimages of
agure painting,a
ower paintingand a
landscapepainting
Image retrieval
technique
173

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT