Experimental Analysis of Digital Image Retrieval Using ... - IEEE Xplore
Experimental Analysis of Digital Image Retrieval Using SVD Deepika Sharma
Pawanesh Abrol
Research Scholar Department of Computer Science and IT University of Jammu, J&K, INDIA [email protected]
Associate Professor Department of Computer Science and IT University of Jammu, J&K, INDIA [email protected]
Abstract – Image Fusion is a technique of combining the useful information from a set of images into a single image, where the output fused image will be more informative and useful than any of the input images. Image fusion techniques can improve the quality and increase the application area of these data. This research paper compares the experimental results generated by the proposed SVD based image retrieval model with the standard method of image retrieval. Keywords – Image Retrieval, SVD, Image fusion, agricultural images
I.
using any of the following methods like discrete wavelet (DWT), stationary wavelet or multi wavelet transform. B. Singular Value Decomposition (SVD) SVD is based on a theorem from linear algebra which says that a rectangular matrix A can be broken down into the product of three matrices - an orthogonal matrix U, a diagonal matrix S, and the transpose of an orthogonal matrix V. The theorem is usually presented something like this: Amn = Umm SmnVTnn
INTRODUCTION
Image fusion can be defined as a process of combining of two images into a single image that has the maximum information [1]. With rapid advancements in technology, it is now possible to obtain information from multi source images to produce a high quality fused image with spatial and spectral information [2] [3]. Image Fusion is a mechanism to improve the quality of information from a set of images. Important applications of the fusion of images include medical imaging, microscopic imaging, remote sensing, and computer vision. A. Image Fusion Techniques The process of digital image fusion involves retaining the good information from each of the given images. This information is fused together to form a resultant image whose quality may be superior to any of the input images. Image fusion methods can be broadly classified into two groups Spatial domain fusion method and Transform domain fusion methods. There are various techniques that have been developed to perform image fusion. Some well-known image fusion methods are listed below [3]:1. Intensity-hue-saturation (IHS) transform 2. Principal component analysis (PCA) 3. Multi scale transform Multi scale transform based image fusion can further be classified as high-pass filtering method, pyramid method wavelet transforms and curvelet transforms. Pyramid method can be Gaussian, laplacian, Gradient, Morphological or low pass pyramid. Wavelet transform, in turn can be performed
c 978-93-80544-12-0/14/$31.00 2014 IEEE
where UTU = I; VTV = I; the columns of U are orthonormal eigenvectors of AAT, the columns of V are orthonormal eigenvectors of ATA, and S is a diagonal matrix containing the square roots of Eigen values from U or V in descending order. SVD has many practical and theoretical values; special features of SVD are that it can be performed on any real (m, n) matrix [4]. II.
LITERAURE REVIEW
Different researchers study various methods and techniques of image fusion of digital images. Various fusion techniques are under consideration which combines the two images with preserving features. We proposed a SVD based image retrieval method from the fused images which gives good results as compared with the image retrieval method using existing method of retrieving. We studied different techniques of image retrieval proposed by different researchers. Various methods investigate effective features and applications of SVD based model. B. Nebel discusses the structure features for content based image retrieval and classification problem in his research study. The image retrieval totally depends on the structure of the image i.e size of the grid [5]. Y. Hui et al. in his research study proposed the various features extracting techniques depending on that the content of the image can be easily retrieved. The main focus of the study is to retrieve only jpeg format images [6]. SVD-based image fusion method for image reconstruction is developed by A. P. Kurian et al. The proposed method shows clear performance improvements compared to the bicubic interpolation and methods of both
911
simulated and real-world images. The advantage of the proposed method lies in its ability to retain the important information in the low resolution images by the SVD-based fusion step and hence improve the image reconstruction results [7]. A new method for automatic indexing and retrieval, implicit higher-order structure from the images is designed by X. S. Zhou et al [8]. In this research work, Singular-value decomposition is used to decompose a large term by document matrix into 50 to 150 orthogonal factors from which the content can be retrieved depending on some specific characteristics. R C. Veltkamp et al. discusses the various features required for the content retrieval process and also designed the technique to extract such features from the images for further processing [9]. R. Datta et al. survey different image retrieval models in order to study the performance of each model [10]. D. Sharma et al. studied the behavior of noise in different digital images using SVD [11]. III. Image Set domains
PROPOSED METHOD
Image used for Analysis of Proposed Model
diagram of the proposed SVD based content retrieval model from real image database. Two different images from any domain of real image database can be input to the model say Image 1 and Image 2. These input images are normalized in the required format for further processing. After normalization SVD is applied to both the images so that the desirable features can be extracted for future use. These images are then combined using some existing method. During combining process, the entire feature of both the images is automatically combined in order to generate a single image which is actually a mixture of two images. After combining, we have to retrieve any image from this combined image. For this, the designed SVD based image retrieval model is applied to this combined image for the desired output image. Id = ∝ . IF + β . Ii + γ
where ∝=
ȕ = 0.4,
0.4,
for gray
level,
Isize
Ȗ = 0.3714
= 50 to 300
∝ , β , γ are constants.
Agricultural
This method is designed for all the real images of size ranging from 50 x 50, 100 x 100,150 x 150, and 300 x 300. Images used in this model are in any format. This proposed model shows effective results for all the images irrespective of any size and format. There are three images in this formula – I1, I2, I3. Using this formula image can be easily retrieved from combined image depending on the features extracted by SVD.
Clouds & planes
Topography
Domain 1
Domain
Fig.1. Standard database image set
Multiple image fusion techniques are applied to a variety of digital images in order to retrieve any image from fused image. However, it is difficult to retrieve back the images from the resultant fused image that are similar to the corresponding input images. The main goal of this research work is to retrieve an image form the input image set that is approximately similar to the corresponding original image. For this a robust SVD based image retrieval method has been developed. For experimental analysis, three standard databases comprising of real images have been used. These consist of images having different sizes i.e. 50 x 50, 100 x 100, 150 x 150 and 300 x 300. These databases consist of various images from different domains like agriculture, clouds position and environment screenshots in order to generate a wide variety of results. Some of the images are shown below in Fig.1.
Image I1
Image I2
Normalization Features of Image I1
Features of Image I2
SVD Feature Extractor
Fusing images by combining features
SVD based Image Retrieval Model
Retrieved Desired Image
Compare Image with corresponding
image
IV. METHODLOGY In this study the efficiency of the proposed SVD based image retrieval model is compared with that of existing standard method. A comparison has been drawn between the two methods of image retrieval. Fig.2. shows the schematic
912
Results
Fig.2. Work Flow of Proposed SVD based IR Model
2014 International Conference on Computing for Sustainable Global Development (INDIACom)
In order to check the efficiency of the proposed model, the retrieved content of the image is compared to the original one. The result in percentage shows the accuracy of the model. More the percentage of matching more is accuracy of the proposed model. This model tests more than 150 images from different domains of real images which enhances the capability of the proposed SVD based image retrieval model.
Graphical representations of the experimental results of different image sizes are shown below (Fig.3.–Fig.6.)
V. RESULTS AND DISCUSSIONS TABLE I. Showing the experimental results for given image database sets
Match percentage of Proposed SVD based IR model & existing model Image sizes Images
50 x 50
100 x100
150 x 150
EM
PM
EM
PM
EM
PM
EM
PM
I1
51.6
73.8
60.7
82.8
67.9
87.0
56.9
89.9
Fig.3. Comparison between the SVD based image retrieval method and image Retrieved using existing method (image Size- 50 x 50)
300 x 300
I2
56.7
78.8
54.8
75.2
69.9
90.8
67.8
88.7
I3
68.8
80.7
58.8
89.9
73.0
89.8
53.0
88.5
I4
59.9
76.3
65.9
71.6
74.8
86.7
57.9
91.7
I5
62.8
81.8
66.8
73.9
58.9
76.9
69.9
83.9
I6
64.9
85.9
69.8
76.9
56.9
83.9
64.5
80.8
I7
61.0
79.7
70.7
82.0
70.7
86.7
45.9
79.8
I8
74.8
85.2
71.9
81.8
50.8
87.5
51.8
78.7
I9
69.9
77.8
65.9
77.9
56.1
89.7
56.9
88.9
I10
77.9
82.9
74.9
89.8
53.0
90.7
49.9
72.8
I11
65.9
76.8
72.9
84.9
51.8
91.8
63.9
75.6
I12
70.9
78.7
52.8
86.6
57.9
85.8
61.9
78.9
I13
79.9
86.8
63.8
79.8
60.7
82.9
62.9
79.6
I14
72.9
81.2
60.8
76.7
61.0
87.7
67.9
84.9
I15
79.8
87.8
56.9
87.9
62.9
87.9
64.1
82.8
Multiple test cases are generated to study the efficiency of the proposed SVD based image retrieval model. The process is carried out in two ways. In the first process, images are combined using standard method and then the existing model is applied to retrieve any image from the fused image and then compared with the corresponding original image. In the second process of proposed study, SVD is applied to the images to combine the two images. After that the proposed SVD based image retrieval model is applied to the fused image in order to retrieve the desired image. The obtained image is then compared with the corresponding original image in order to check the efficiency of the model. The retrieved image seems to be approximately similar with that of the corresponding original image. Thus the proposed model is more efficient in image retrieving process as compared to the existing model of image retrieval. More than 150 images from different databases have been taken to investigate the extent of proposed SVD based image retrieving model. A comparison has been given on the basis of these results showing the efficiency of the proposed SVD based image retrieval model with that of the image retrieved using the standard model.
Fig.4. Comparison between the SVD based Image retrieval Method & Image Retrieved using existing method (image Size-100 x 100)
Fig.5. Comparison between the SVD based image retrieval method & image retrieved using existing method (image Size-150 x 150)
Fig.6. Comparison between the SVD based image retrieval method & image retrieved using existing method (Image Size-300 x 300)
2014 International Conference on Computing for Sustainable Global Development (INDIACom)
913
VI. CONCLUSION AND FUTURE WORK The experimental analysis indicates that the proposed IR SVD model can provide a significant feature based alternative method to detect and prevent different kinds of digital image tampering including copy move. The proposed SVD based image retrieval model gives the satisfactory results in varied domains and sizes of real images. This model extracts the features from the fused images. Depending upon the feature vectors the input image can be retrieved more effective. The proposed model can be used for prevention of image tempering and enhancing security. The model can further be explored for network based image streams and for more complex and real time images. The method of retrieving the images shall be extended or refined for multiple features of the images.
REFERENCES [1]
D. A.Godse, and D. S. Bormane, “Wavelet based image fusion using pixel based maximum selection rule” International Journal of Engineering Science and Technology (IJEST), Vol. 3 No. 7, July 2011. [2] S. Vekkot, and P. Shukla, “A Novel Architecture for Wavelet based Image Fusion”, World Academy of Science, Engineering and Technology, 2009. [3] S. Huang, G. Gul, I. Avcibas, and F. Kurugollu, “Wavelet for Image Fusion”, in Proc. 17th Int. Conf. on Image Processing and digital signal technology, pp. 1765-1768, 2010. [4] O. N. Osmanli, “A Singular Value Decomposition Approach for Recommendations Systems”, M.Sc. thesis, Dept. of Computer Engineering, Middle East Technical University, 2010. [5] B.Nebel, “Structure Features for Content-Based Image Retrieval and Classification Problems”, M.Sc. Dissertation, 2007. [6] Y.Hui, L.Mingjing, Z. Hong-Jiang and F. Jufu, “Color texture moments for content-based image retrieval”, in Proc. International Conference on Image Processing, vol. 3 pp. 929 - 932, 2002. [7] A. P. Kurian, S. R, Lekshmi Mohan, M. M. Kartha and K. P. Soma, “SVD based image manipulations detection”, International Journal of Computer Applications”, vol.58, no. 12, 2012. [8] X.S.Zhou & T. S. Huang, “Relevance feedback in image retrieval: a comprehensive review”, Int. Journal of Multimedia Systems, vol. 8, 2003. [9] R. C. Veltkamp, J.Schietse and J.P. Eakins, “Practice and Challenges in Trademark Image Retrieval”, Int. Journal of Image processing and computer vision, vol.6, no.1, 2007. [10] R. Datta, D. Joshi, J. Li, and J. Z. Wang, “Image Retrieval: Ideas, Influences, and Trends of the New Age”, ACM Computing Surveys, Vol. 40, No. 2, Article 5, 2008. [11] D. Sharma and P. Abrol, “Investigating the Extent of Noise in Digital Images using singular value decomposition”, Int J. of Software and Web Services (IJSWS), vol.1, no. 4, pp. 6-14, Mar – May, 2013 (code IJSWS 13 111) (ISSN 22790063 (Print), 2279-0071 (online)) E-version available: http://iasir.net/ijswsissue/ijswsis sue 4-1.html
914
2014 International Conference on Computing for Sustainable Global Development (INDIACom)
Pawanesh Abrol
Research Scholar Department of Computer Science and IT University of Jammu, J&K, INDIA [email protected]
Associate Professor Department of Computer Science and IT University of Jammu, J&K, INDIA [email protected]
Abstract – Image Fusion is a technique of combining the useful information from a set of images into a single image, where the output fused image will be more informative and useful than any of the input images. Image fusion techniques can improve the quality and increase the application area of these data. This research paper compares the experimental results generated by the proposed SVD based image retrieval model with the standard method of image retrieval. Keywords – Image Retrieval, SVD, Image fusion, agricultural images
I.
using any of the following methods like discrete wavelet (DWT), stationary wavelet or multi wavelet transform. B. Singular Value Decomposition (SVD) SVD is based on a theorem from linear algebra which says that a rectangular matrix A can be broken down into the product of three matrices - an orthogonal matrix U, a diagonal matrix S, and the transpose of an orthogonal matrix V. The theorem is usually presented something like this: Amn = Umm SmnVTnn
INTRODUCTION
Image fusion can be defined as a process of combining of two images into a single image that has the maximum information [1]. With rapid advancements in technology, it is now possible to obtain information from multi source images to produce a high quality fused image with spatial and spectral information [2] [3]. Image Fusion is a mechanism to improve the quality of information from a set of images. Important applications of the fusion of images include medical imaging, microscopic imaging, remote sensing, and computer vision. A. Image Fusion Techniques The process of digital image fusion involves retaining the good information from each of the given images. This information is fused together to form a resultant image whose quality may be superior to any of the input images. Image fusion methods can be broadly classified into two groups Spatial domain fusion method and Transform domain fusion methods. There are various techniques that have been developed to perform image fusion. Some well-known image fusion methods are listed below [3]:1. Intensity-hue-saturation (IHS) transform 2. Principal component analysis (PCA) 3. Multi scale transform Multi scale transform based image fusion can further be classified as high-pass filtering method, pyramid method wavelet transforms and curvelet transforms. Pyramid method can be Gaussian, laplacian, Gradient, Morphological or low pass pyramid. Wavelet transform, in turn can be performed
c 978-93-80544-12-0/14/$31.00 2014 IEEE
where UTU = I; VTV = I; the columns of U are orthonormal eigenvectors of AAT, the columns of V are orthonormal eigenvectors of ATA, and S is a diagonal matrix containing the square roots of Eigen values from U or V in descending order. SVD has many practical and theoretical values; special features of SVD are that it can be performed on any real (m, n) matrix [4]. II.
LITERAURE REVIEW
Different researchers study various methods and techniques of image fusion of digital images. Various fusion techniques are under consideration which combines the two images with preserving features. We proposed a SVD based image retrieval method from the fused images which gives good results as compared with the image retrieval method using existing method of retrieving. We studied different techniques of image retrieval proposed by different researchers. Various methods investigate effective features and applications of SVD based model. B. Nebel discusses the structure features for content based image retrieval and classification problem in his research study. The image retrieval totally depends on the structure of the image i.e size of the grid [5]. Y. Hui et al. in his research study proposed the various features extracting techniques depending on that the content of the image can be easily retrieved. The main focus of the study is to retrieve only jpeg format images [6]. SVD-based image fusion method for image reconstruction is developed by A. P. Kurian et al. The proposed method shows clear performance improvements compared to the bicubic interpolation and methods of both
911
simulated and real-world images. The advantage of the proposed method lies in its ability to retain the important information in the low resolution images by the SVD-based fusion step and hence improve the image reconstruction results [7]. A new method for automatic indexing and retrieval, implicit higher-order structure from the images is designed by X. S. Zhou et al [8]. In this research work, Singular-value decomposition is used to decompose a large term by document matrix into 50 to 150 orthogonal factors from which the content can be retrieved depending on some specific characteristics. R C. Veltkamp et al. discusses the various features required for the content retrieval process and also designed the technique to extract such features from the images for further processing [9]. R. Datta et al. survey different image retrieval models in order to study the performance of each model [10]. D. Sharma et al. studied the behavior of noise in different digital images using SVD [11]. III. Image Set domains
PROPOSED METHOD
Image used for Analysis of Proposed Model
diagram of the proposed SVD based content retrieval model from real image database. Two different images from any domain of real image database can be input to the model say Image 1 and Image 2. These input images are normalized in the required format for further processing. After normalization SVD is applied to both the images so that the desirable features can be extracted for future use. These images are then combined using some existing method. During combining process, the entire feature of both the images is automatically combined in order to generate a single image which is actually a mixture of two images. After combining, we have to retrieve any image from this combined image. For this, the designed SVD based image retrieval model is applied to this combined image for the desired output image. Id = ∝ . IF + β . Ii + γ
where ∝=
ȕ = 0.4,
0.4,
for gray
level,
Isize
Ȗ = 0.3714
= 50 to 300
∝ , β , γ are constants.
Agricultural
This method is designed for all the real images of size ranging from 50 x 50, 100 x 100,150 x 150, and 300 x 300. Images used in this model are in any format. This proposed model shows effective results for all the images irrespective of any size and format. There are three images in this formula – I1, I2, I3. Using this formula image can be easily retrieved from combined image depending on the features extracted by SVD.
Clouds & planes
Topography
Domain 1
Domain
Fig.1. Standard database image set
Multiple image fusion techniques are applied to a variety of digital images in order to retrieve any image from fused image. However, it is difficult to retrieve back the images from the resultant fused image that are similar to the corresponding input images. The main goal of this research work is to retrieve an image form the input image set that is approximately similar to the corresponding original image. For this a robust SVD based image retrieval method has been developed. For experimental analysis, three standard databases comprising of real images have been used. These consist of images having different sizes i.e. 50 x 50, 100 x 100, 150 x 150 and 300 x 300. These databases consist of various images from different domains like agriculture, clouds position and environment screenshots in order to generate a wide variety of results. Some of the images are shown below in Fig.1.
Image I1
Image I2
Normalization Features of Image I1
Features of Image I2
SVD Feature Extractor
Fusing images by combining features
SVD based Image Retrieval Model
Retrieved Desired Image
Compare Image with corresponding
image
IV. METHODLOGY In this study the efficiency of the proposed SVD based image retrieval model is compared with that of existing standard method. A comparison has been drawn between the two methods of image retrieval. Fig.2. shows the schematic
912
Results
Fig.2. Work Flow of Proposed SVD based IR Model
2014 International Conference on Computing for Sustainable Global Development (INDIACom)
In order to check the efficiency of the proposed model, the retrieved content of the image is compared to the original one. The result in percentage shows the accuracy of the model. More the percentage of matching more is accuracy of the proposed model. This model tests more than 150 images from different domains of real images which enhances the capability of the proposed SVD based image retrieval model.
Graphical representations of the experimental results of different image sizes are shown below (Fig.3.–Fig.6.)
V. RESULTS AND DISCUSSIONS TABLE I. Showing the experimental results for given image database sets
Match percentage of Proposed SVD based IR model & existing model Image sizes Images
50 x 50
100 x100
150 x 150
EM
PM
EM
PM
EM
PM
EM
PM
I1
51.6
73.8
60.7
82.8
67.9
87.0
56.9
89.9
Fig.3. Comparison between the SVD based image retrieval method and image Retrieved using existing method (image Size- 50 x 50)
300 x 300
I2
56.7
78.8
54.8
75.2
69.9
90.8
67.8
88.7
I3
68.8
80.7
58.8
89.9
73.0
89.8
53.0
88.5
I4
59.9
76.3
65.9
71.6
74.8
86.7
57.9
91.7
I5
62.8
81.8
66.8
73.9
58.9
76.9
69.9
83.9
I6
64.9
85.9
69.8
76.9
56.9
83.9
64.5
80.8
I7
61.0
79.7
70.7
82.0
70.7
86.7
45.9
79.8
I8
74.8
85.2
71.9
81.8
50.8
87.5
51.8
78.7
I9
69.9
77.8
65.9
77.9
56.1
89.7
56.9
88.9
I10
77.9
82.9
74.9
89.8
53.0
90.7
49.9
72.8
I11
65.9
76.8
72.9
84.9
51.8
91.8
63.9
75.6
I12
70.9
78.7
52.8
86.6
57.9
85.8
61.9
78.9
I13
79.9
86.8
63.8
79.8
60.7
82.9
62.9
79.6
I14
72.9
81.2
60.8
76.7
61.0
87.7
67.9
84.9
I15
79.8
87.8
56.9
87.9
62.9
87.9
64.1
82.8
Multiple test cases are generated to study the efficiency of the proposed SVD based image retrieval model. The process is carried out in two ways. In the first process, images are combined using standard method and then the existing model is applied to retrieve any image from the fused image and then compared with the corresponding original image. In the second process of proposed study, SVD is applied to the images to combine the two images. After that the proposed SVD based image retrieval model is applied to the fused image in order to retrieve the desired image. The obtained image is then compared with the corresponding original image in order to check the efficiency of the model. The retrieved image seems to be approximately similar with that of the corresponding original image. Thus the proposed model is more efficient in image retrieving process as compared to the existing model of image retrieval. More than 150 images from different databases have been taken to investigate the extent of proposed SVD based image retrieving model. A comparison has been given on the basis of these results showing the efficiency of the proposed SVD based image retrieval model with that of the image retrieved using the standard model.
Fig.4. Comparison between the SVD based Image retrieval Method & Image Retrieved using existing method (image Size-100 x 100)
Fig.5. Comparison between the SVD based image retrieval method & image retrieved using existing method (image Size-150 x 150)
Fig.6. Comparison between the SVD based image retrieval method & image retrieved using existing method (Image Size-300 x 300)
2014 International Conference on Computing for Sustainable Global Development (INDIACom)
913
VI. CONCLUSION AND FUTURE WORK The experimental analysis indicates that the proposed IR SVD model can provide a significant feature based alternative method to detect and prevent different kinds of digital image tampering including copy move. The proposed SVD based image retrieval model gives the satisfactory results in varied domains and sizes of real images. This model extracts the features from the fused images. Depending upon the feature vectors the input image can be retrieved more effective. The proposed model can be used for prevention of image tempering and enhancing security. The model can further be explored for network based image streams and for more complex and real time images. The method of retrieving the images shall be extended or refined for multiple features of the images.
REFERENCES [1]
D. A.Godse, and D. S. Bormane, “Wavelet based image fusion using pixel based maximum selection rule” International Journal of Engineering Science and Technology (IJEST), Vol. 3 No. 7, July 2011. [2] S. Vekkot, and P. Shukla, “A Novel Architecture for Wavelet based Image Fusion”, World Academy of Science, Engineering and Technology, 2009. [3] S. Huang, G. Gul, I. Avcibas, and F. Kurugollu, “Wavelet for Image Fusion”, in Proc. 17th Int. Conf. on Image Processing and digital signal technology, pp. 1765-1768, 2010. [4] O. N. Osmanli, “A Singular Value Decomposition Approach for Recommendations Systems”, M.Sc. thesis, Dept. of Computer Engineering, Middle East Technical University, 2010. [5] B.Nebel, “Structure Features for Content-Based Image Retrieval and Classification Problems”, M.Sc. Dissertation, 2007. [6] Y.Hui, L.Mingjing, Z. Hong-Jiang and F. Jufu, “Color texture moments for content-based image retrieval”, in Proc. International Conference on Image Processing, vol. 3 pp. 929 - 932, 2002. [7] A. P. Kurian, S. R, Lekshmi Mohan, M. M. Kartha and K. P. Soma, “SVD based image manipulations detection”, International Journal of Computer Applications”, vol.58, no. 12, 2012. [8] X.S.Zhou & T. S. Huang, “Relevance feedback in image retrieval: a comprehensive review”, Int. Journal of Multimedia Systems, vol. 8, 2003. [9] R. C. Veltkamp, J.Schietse and J.P. Eakins, “Practice and Challenges in Trademark Image Retrieval”, Int. Journal of Image processing and computer vision, vol.6, no.1, 2007. [10] R. Datta, D. Joshi, J. Li, and J. Z. Wang, “Image Retrieval: Ideas, Influences, and Trends of the New Age”, ACM Computing Surveys, Vol. 40, No. 2, Article 5, 2008. [11] D. Sharma and P. Abrol, “Investigating the Extent of Noise in Digital Images using singular value decomposition”, Int J. of Software and Web Services (IJSWS), vol.1, no. 4, pp. 6-14, Mar – May, 2013 (code IJSWS 13 111) (ISSN 22790063 (Print), 2279-0071 (online)) E-version available: http://iasir.net/ijswsissue/ijswsis sue 4-1.html
914
2014 International Conference on Computing for Sustainable Global Development (INDIACom)
