Figures
Abstract
A complete texture image retrieval system includes two techniques: texture feature extraction and similarity measurement. Specifically, similarity measurement is a key problem for texture image retrieval study. In this paper, we present an effective similarity measurement formula. The MIT vision texture database, the Brodatz texture database, and the Outex texture database were used to verify the retrieval performance of the proposed similarity measurement method. Dual-tree complex wavelet transform and nonsubsampled contourlet transform were used to extract texture features. Experimental results show that the proposed similarity measurement method achieves better retrieval performance than some existing similarity measurement methods.
Citation: Zhu Z, Zhao C, Hou Y (2012) Research on Similarity Measurement for Texture Image Retrieval. PLoS ONE 7(9): e45302. https://doi.org/10.1371/journal.pone.0045302
Editor: Helmut Ahammer, Medical University of Graz, Austria
Received: April 25, 2012; Accepted: August 20, 2012; Published: September 25, 2012
Copyright: © Zhu et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the National Science Foundation of China under Grants 90820306 and 61072148. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
With the rapid expansion of digital image libraries and multimedia databases, content-based image retrieval (CBIR) has become a hot research topic in the computer science field. Content of images includes color characteristics, shape characteristics, texture characteristics, and semantics characteristics. Because texture is a type of inherent property for most physical surfaces, texture characteristics usually play an important role in CBIR.
A complete texture image retrieval system includes two techniques: texture features extraction and similarity measurement. Texture features used in CBIR are usually extracted by space-frequency domain approaches [1]–[8]. Shutao Li and John Shawe-Taylor used a wavelet transform and a contourlet transform to extract texture features for image classifying [9]. A Gabor filter has a nice effect on descripting texture features, but it still has two shortcomings: one is that redundant information is produced after different Gabor filters; the other is that feature extraction by Gabor filters usually has considerably high computational complexity. N. Kingsbury et al used a dual-tree complex wavelet transform (DT-CWT) to extract texture features [10]–[12]. DT-CWT can overcome two drawbacks of the Discrete Wavelet Transform (DWT); one is that invariance, the other is that DWT has only limited directivity. DT-CWT not only has good localization in time-frequency domains, but also has approximate translation invariance, more directivity and limited data redundancy. A nonsubsampled contourlet transform (NSCT) has anisotropy and translation invariance [13]–[14]. In this paper, DT-CWT and NSCT are respectively used to extract texture features of images.
Similarity measurement is a key technique for texture image retrieval. Kokare. M. et al in [15] compared nine distance similarity measurements, such as Weighted-Mean-Variance distance (WMVD), Euclidean distance (ED), Canberra distance (CD), Bray-Curtis distance (BCD), Manhattan distance, Mahalanobis distance, Chebyshev distance, Squared Chi-Squared distance, and Squared Chord distance for texture image retrieval. Experimental results show that WMVD, CD and BCD are the three best distance similarity measurements for image retrieval problems, but the retrieval rates are not ideal when using WMVD, CD and BCD. Therefore, exploring more effective similarity measurements is a problem worth studying.
In this paper, we present an effective similarity measurement. A dual-tree complex wavelet transform and a nonsubsampled contourlet transform were respectively used to extract texture features. The MIT vision texture database (640 images), the Brodatz texture database (1776 images), and the Outex texture database (5104 images) were used to verify the retrieval performance. Experimental results show that the retrieval performance can be improved by the proposed similarity measurement more than some existing similarity distance measurement methods.
Methods
2.1 Related Works
Because WMVD, CD and BCD are the best three similarity measurements for image retrieval [15], so here we focus on the WMVD similarity measurement, the CD similarity measurement and the BCD similarity measurement.
WMVD (Weighted-Mean-Variance distance) is widely used in image retrieval [1] [7]. Generally, two patterns and
are considered, where
is a query image and
is a target image in the database.
and
are, respectively, the feature vectors of
and
. The WMV is defined as the following,
(1)
WhereWhere
denotes the scale, and
is the number of subbands in each scale.
and
are the mean and the standard deviation of each subband for a query image.
and
are the mean and the standard deviation of each subband for a target image.
and
are the standard deviations of
and
respectively over the entire database, and they are used to normalize the individual feature components.
If and
are two n-dimensional feature vectors of an image to retrieve and query the image:
The ED is defined as the following,(2)the CD is defined as the following,
(3)and the BCD is defined as the following,
(4)In this paper, we present an effective distance measurement that is more effective than the above three distance measurement methods.
(a) NSFB structure that implements the NSCT (b) The obtained Frequency partitioning.
2.2 Proposed Distance Similarity Measure
We present an effective Average Euclidean distance (AED). If X and Y are two n-dimensional feature vectors of an image from a database and the query image, we give the new similarity measurement as the following,(5)
The proposed distance similarity measurement not only contains some relations between objects, but also comprehensively considers all dimensional feature parameters. When the proposed similarity distance measurement is used for texture image retrieval, the experimental results show that the retrieval performances are better using the proposed similarity distance measurement than the existing similarity measurements.
2.3 Extraction of Texture Features
2.3.1 Extraction of texture features based on nonsubsampled contourlet transform.
A nonsubsampled contourlet transform includes a nonsubsampled pyramid and nonsubsampled directional filter banks [13]–[14]. A nonsubsampled pyramid includes a set of two-channel nonsubsampled filter banks (NSFB). Nonsubsampled filtering does not implement a downsampling operation on an image but implements upsampling for filter banks, so NSCT has not only anisotropy but also the shift invariance.
Two–level NSCT decomposition is shown in Figure 1.
A nonsubsampled Laplace pyramid is a two-channel nonsubsampled transform. The condition of perfect reconstruction is shown as the following,(6)Where,
and
denote low frequency and high frequency decomposition filters.
and
denote low frequency and high-frequency reconstruction filters. For practical image decomposition, a nonsubsampled àtrous wavelet transform is used to obtain a high frequency subband and a low frequency subband; a certain number of directional filters are then used to get some directional subbands. In order to get multiresolution analysis, one can continue to decompose the à trous wavelet transform low frequency subband.
Each image is decomposed five levels using a nonsubsampled Laplace pyramid decomposition; all the obtained high frequency subbands then continue to be decomposed by the nonsubsampled directional filter banks. A “pyr” pyramidal filter and a “vk” directional filter are used to lower the time complexity in our experiments, because they both have relatively small support. After the above decomposition process, 31 high frequency subbands and 1 low frequency subband are obtained. The mean and the standard deviation
of the subband coefficients are calculated. There are 32 subbands: the feature vector
is constructed by
and
as the following,
(7)
2.3.2 Extraction of texture features based on a dual-Tree complex wavelet transform.
DT-CWT is usually used to extract texture features in a wavelet domain. An image can be decomposed into two low frequency subbands and six high frequency subbands by DT-CWT in every level. The low frequency subbands can be decomposed again. In our experiments, an image is decomposed into three levels by DT-CWT. All together, there are two low frequency subbands and eighteen high frequency subbands. The mean and the standard deviation
of the coefficients in each subband are calculated. There are 20 subbands; the feature vector
is constructed by
and
as the following,
(8)
The MIT image database (640 images) were used.
The MIT image database (640 images) were used.
The Brodatz image database (1776 images) were used.
The Brodatz image database (1776 images) were used.
The Outex image database (5104 images) were used.
The Outex image database (5104 images) were used.
Results
3.1 Image Database
1) MIT image database.
In the retrieval experiments, real-world images of different natural scenes from the Massachusetts Institute of Technology (MIT) Vision Texture database are used [16]. There are 640 texture images from 40 different classes from the MIT texture database in the image database. Each original MIT texture image is divided into sixteen nonoverlapping
subimages. The total number of images in this database is 640 (
). The query image is any one of the 640 subimages; the other 15 subimages from the same class are relevant candidate images. Forty different classes from the MIT texture database are shown as Figure 2.
2) Brodatz image database.
There are 1776 texture images from 111 different classes from the Brodatz texture database [17] in the image database. Each original Brodatz texture image is divided into sixteen nonoverlapping subimages. The total number of images in this database is 1776 (
). The query image is any one of the 1776 subimages; the other 15 subimages from the same class are relevant candidate images. Different classes from the Brodatz texture database are shown as Figure 3.
3) Outex image database.
There are 5104 texture images from 319 different classes from the Outex texture database [18] in the image database. Each original Outex texture image is divided into sixteen nonoverlapping subimages. The total number of images in this database is 5104 (
). The query image is any one of the 5104 subimages; the other 15 subimages from the same class are relevant candidate images. Different classes from the Outex texture database are shown as Figure 4.
3.2 The Existing and the Proposed Texture Image Retrieval Methods
In our experiments, each image is decomposed by DT-CWT and NSCT respectively. Ten kinds of texture image retrieval methods are given as follows:
Method 1: use DT-CWT and ED.
Method 2: use DT-CWT and WMVD.
Method 3: use DT-CWT and CD.
Method 4: use DT-CWT and BCD.
Method 5 : use DT-CWT and AED.
Method 6: use NSCT and ED.
Method 7: use NSCT and WMVD.
Method 8: use NSCT and CD.
Method 9: use NSCT and BCD.
Method 10 : use NSCT and AED.
3.3 Experimental Results
The average precision ratio is used to evaluate retrieval performance. The average precision ratio is calculated using the following formula,(9)Where
is the total number of the images in the texture database.
is the number of similar images that belong to the same class in the image database.
is the number of images that are properly ferreted out from the texture database in practice. In this paper, there are sixteen
subimages in the same class. The number of top retrieved images is considered as 16.
is the number of images of the top 16 retrieved images belonging to the same class.
The proposed method improves retrieval performance on the database, compared with the other four similarity measurements. The experimental results are shown as Table 1, Table 2, Table 3, Table 4, Table 5, and Table 6.
We evaluate the performance in terms of the average retrieval rate of relevant images as a function of the number of top retrieved images; the retrieval performance is shown as Figure 5, Figure 6, Figure 7, Figure 8, Figure 9, and Figure 10. The experimental results show that the proposed similarity measurement can improve average precision on the image retrieval.
Discussion
The ED measurement lacks the interrelations between objects. When the ED measurement is used as a similarity measurement for texture image retrieval in a wavelet domain, retrieval accuracy is not always satisfied. The WMVD similarity measurement has been widely used for image retrieval, but it has a shortcoming: the similarity measurement between two images is sometimes affected by some uncorrelated images in the whole image database, because the standard deviation of all the images in the whole database needs to be calculated in order to normalize the Euclidean distance. CD and BCD similarity measurements use the difference and the normalization of the difference between two image features; they do not have scaling effects, but their drawback is that they both sum differences in a simple manner. The proposed similarity measurement first uses the denominator to normalize the difference between the two image features, so it can avoid scaling effects. Next, the differences in each dimension are squared, and they are summed before extracting the square root. The proposed similarity measurement can comprehensively use all features; it can avoid the limitations of CD and BCD similarity measurements.
Conclusion
We present an effective similarity distance measurement for image retrieval. Features of all the images were extracted using DT-CWT and NSCT respectively. Experimental results demonstrate that the proposed similarity distance measurement achieves higher retrieval accuracy than some existing similarity measures.
Acknowledgments
The authors would like to thank all the anonymous reviewers for their valuable comments.
Author Contributions
Conceived and designed the experiments: ZZ CZ YH. Performed the experiments: ZZ CZ. Analyzed the data: ZZ YH. Contributed reagents/materials/analysis tools: ZZ CZ. Wrote the paper: ZZ CZ YH.
References
- 1. Manjunath BS, Ma WY (1996) Texture features for browsing and retrieval of image data, IEEE Trans. IEEE Transactions on Pattern Analysis and Machine Intelligence 18(8): 837–842.
- 2. Laine A, Fan J (1993) Texture classification by wavelet packet signatures. IEEE Transactions on Pattern Analysis and Machine Intelligence 15(11): 1186–1191.
- 3. Unser M (1995) Texture classification and segmention using wavelet frames. IEEE Trans. IEEE Transactions on Image Processing 4(11): 1549–1560.
- 4. Wouwer GV, Scheunders P, Dyck DV (1999) Statistical texture characterization from discrete wavelet representations. IEEE Transactions on Image Processing 8(4): 592–598.
- 5. Randen T, Husoy J (1999) Filtering for texture classification: A comparative study. IEEE Transactions on Pattern Analysis and Machine Intelligence 21(4): 291–310.
- 6. Do M, Vetterli M (2002) Wavelet-based texture retrieval using generalized Gassian density and Kullback-Leibler distance. IEEE Transactions on Image Processing 11(2): 146–158.
- 7. Kokare M, Biswas PK, Chatterji BN (2005) Texture image retrieval using new rotated complex wavelet filters. IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics 35(6): 1168–1178.
- 8. Ves E, Ruedin A, Acevedo D, Benavent C, Seijas L (2007) A new wavelet-based texture descriptor for image retrieval. Proceedings of the 12th International Conference on Computer Analysis of Images and Patterns 4673: 895–902.
- 9. Shutao L, John ST (2004) Texture classification by combing wavelet and contourlet features. Joint IAPR International workshops on Structural and Syntactical Pattern Recognition and Statical Pattern Recognition 3138: 1126–1134.
- 10. Kingsbury N (1998) The dual-tree complex wavelet transform: A new efficient tool for image restoration and enhancement. Proceeding of European Signal Processing Conference 1: 319–322.
- 11. Rivaz P, Kingsbury N (1999) Complex wavelet features for fast texture image retrieval. Proceedings of the IEEE Conference on Image Processing 1: 109–113.
- 12. Selesnick IW, Baraniuk RG, Kingsbury NG (2005) The dual-tree complex wavelet transform. IEEE Transactions on Signal Processing 22(6): 123–151.
- 13. Zhou JP, Cunha AL, Do MN (2005) The Nonsubsampled Contourlet Transform: construction and Application in enhancement. IEEE International Conference on Image Processing 1: 469–472.
- 14. Cunha AL, Zhou JP, Do MN (2006) The Nonsubsampled Contourlet Transform: Theory, Design, and Applications. IEEE Transactions on Image Processing 15(10): 3089–3101.
- 15. Kokare M, Chatterji BN, Biswas PK (2003) Comparison of similarity metrics for texture image retrieval. IEEE International Conference on Convergent Technologies for the Asia-Pacific Region 2: 571–575.
- 16.
MIT Vision and Modeling Group. Vision Texture. Available: http://vismod.www.media.mit.edu. Accessed 2012 Jan 8.
- 17.
Brodatz P (1996) Textures: A Photographic Album for Artists and Designers. Available: http://www.ux.uis.no/tranden/~brodatz.html. Accessed 2012 Jun 5.
- 18.
Outex website. Available: http://www.outex.oulu.fi/. Accessed 2012 Jun 6.