International Journal of Application or Innovation in Engineering & Management (IJAIEM)
Web Site: www.ijaiem.org Email: editor@ijaiem.org, editorijaiem@gmail.com
Volume 1, Issue 2, October 2012 ISSN 2319 - 4847
1
2
1
M.E. student at SSCET Bhilai (C.G.) INDIA
2
Associate Professor ETC department, SSCET Bhilai (C.G.) INDIA
BSTRACT
The focus of this work is on using transformation technique for searching, browsing and retrieving images from a large database. In this paper we are using Discrete Cosine Transform to generate the feature vector .Here we are considering DCT of each R, G and B components of an image. The proposed work experimented over a database of 800 images spread across 8 classes of images. Euclidian distances between the feature vectors of query image and the database images are considered.
Images are retrieved in ascending order of Euclidian distances. The Average precision and Average recall of each class and overall average of all averages of each class are calculated as a performance measure .
Keywords: Transformation technique, Feature vector, discrete cosine transform, precision, recall
NTRODUCTION
The large numbers of image collections have posed increasing technical challenges to computer systems to store/transmit and index/manage image data effectively. With advances in digital technology there is an exponential increase in image database in almost all areas. To improve human access, however, there must be an effective and precise method for users to search, browse, and interact with these collections and do so in a timely manner. [1]
Content-based image retrieval, also known as query by image content and content-based visual information retrieval is the application of computer vision to the image retrieval problem, that is, the problem of searching for digital images in large databases. In the CBIR system, the relevance between a query and any target image is ranked according to a similarity measure computed from the visual features. In CBIR each image that is stored in the database has its features extracted and compared to the features of the query image. It involves two steps:
• Feature Extraction: The first step in this process is to extract the image features to a distinguishable extent. Feature is anything that is localized, meaningful and detectable. In an image noticeable features include corners, lines, objects, color, shape, spatial location, motion and texture.
• Matching: The second step involves matching these features to yield a result that is visually similar. [3]
The accuracy of a CBIR system depends on various factors like extraction of appropriate features from the image, classification of the image. The features extracted from an image are represented in the form of vectors. The retrieval accuracy, computational complexity, retrieval time depend on the dimension of the feature vector [7]. Higher the dimension of feature vectors, better the retrieval accuracy but the memory for storage, retrieval time and computational complexity increases. Thus is it important to find balance between the dimension of feature vectors and the accuracy with satisfactory storage and computational requirements.
In this paper an approach is proposed to minimize the size of feature vector to reduce the system complexity and to obtain faster retrieval .In this paper it is proposed to extract the feature vector using frequency domain analysis of the image [10]. Here we obtain the transform coefficients of applied transform and use them as image feature. Frequency domain calculations were made using DCT.
Figure 1 : Architecture of CBIR

International Journal of Application or Innovation in Engineering & Management (IJAIEM)
Web Site: www.ijaiem.org Email: editor@ijaiem.org, editorijaiem@gmail.com
Volume 1, Issue 2, October 2012 ISSN 2319 - 4847
Like any Fourier-related transform, discrete cosine transforms (DCTs) express a function or a signal in terms of a sum of sinusoids with different frequencies and amplitudes. Like the discrete Fourier transforms (DFT), a DCT operates on a function at a finite number of discrete data points. The obvious distinction between a DCT and a DFT is that the former uses only cosine functions, while the latter uses both cosines and sines (in the form of complex exponentials). It is a separable linear transformation; that is, the two-dimensional transform is equivalent to a one dimensional DCT performed along a single dimension followed by a one-dimensional DCT in the other dimension
F[m,n]= -------- (1)
Where C(m),C(n)=1/ for m ,n=0 and C(m),C(n)=1 otherwise .
For full 2-Dimensional DCT for an N x N image the number of multiplications required are N
2
(2N) and number of additions required are N
2
(2N-2).
The DCT coefficients of an image tend themselves as a new feature, which have the ability to represent the regularity, complexity and some texture features of an image and it can be directly applied to image data in the compressed domain [12] This may be a way to solve the large storage space problem and the computational complexity of the existing methods.
The method of image search and retrieval proposed here mainly focuses on the generation of the feature vector of search based on the image transform generated by the discrete cosine transform (DCT).
Steps of the algorithm are given below.
Step1: Extract Red, Green and Blue components of the color image .
Step2: Apply the Transform DCT on individual color planes of image to extract feature vector.
.
Step3: The left topmost component (DC term) of DCT of each color components i.e. R, G and B are taken as
a image feature and stored in feature vector.
Step4: The Euclidian distances between the feature vectors of query image and the feature vectors of images in
the database are calculated.
Step5: The algorithm performance is measured based on the average precision and average recall of each class of
images and their average across the class.
4.1 Development Environment
The functional code for our prototype system was implemented using MATLAB ® 7.0 on a Pentium dual core II, 2.48
GHz Windows-based PC.
4.2 Database Preparation
The database consists of different categories such as Africans and villages, Beaches, Buildings, Buses, Dinosaurs,
Elephants, Flowers, Horses, Mountains and glaciers, Food and Natural scenes collected from the open source [8]. All the categories are used for retrieval. These images are stored in JPEG format with size 384x256 and 256×384, each image is represented with RGB color space.
4.3 Determining Feature Vector
A color image is a combination of some basic colors. In MATLAB we can break each individual pixel of a color image
(termed ‘true color’) down into Red, Green and Blue values as shown in figure 2.
In this method first we divide the color image into its R, G, B components, and then DCT is applied to obtain DCT coefficients.
The first coefficient (DC component) represents the energy information. The first coefficient (DC component) of each transformed color component i.e. R, G & B is used for feature vector generation, so in this method for each color component we obtain one feature value. the feature vector of one RGB image consist of three feature values one feature corresponding to each color component . Once the feature vectors are generated for all images in the database, they are stored in a feature database
4.4 Similarity measure
For similarity comparison, we have used Euclidean distance d using equation d =
2 ---------- (2)
Where Fq[i] is the i th query image feature, and Fdb[i] is the corresponding feature in the feature vector database.

International Journal of Application or Innovation in Engineering & Management (IJAIEM)
Web Site: www.ijaiem.org Email: editor@ijaiem.org, editorijaiem@gmail.com
Volume 1, Issue 2, October 2012 ISSN 2319 - 4847
Figure 2 : Flow diagram of feature vector generation
Database of 800 images of 8 different classes is used to check the performance of the algorithm developed. In order to measure retrieval effectiveness for an image retrieval system, precision and recall values are used. Five different images of each category are used as query images, some representative sample images which are used as query images are shown in Figure 3.
Figure 3 : Sample query Images
Once the feature vectors are generated for all images in the database, they are stored in a feature database. A feature vector of query image of each class is calculated to search the feature database. The sorted Euclidian distance in ascending order between the query image and the database images feature vectors are used to calculate the precision and recall to measure the retrieval performance of the algorithm as shown in the equations (3) and (4).
Precision = ---------- (3)
Recall = ----------- (4)
Figure 4 : top 10 retrieved images for horse class

International Journal of Application or Innovation in Engineering & Management (IJAIEM)
Web Site: www.ijaiem.org Email: editor@ijaiem.org, editorijaiem@gmail.com
Volume 1, Issue 2, October 2012 ISSN 2319 - 4847
Figure 5 : top 10 retrieved images for flower class
For figure 4 Horse and for figure 5 flower is taken as query image to search the given database. The algorithm applied to generate feature vector for each image in the database and the query image. This algorithm has produced good results as it can be seen in the Figure 4 & 5 where 10 retrieved images are of same category as the query image. The first image is the query image itself if it is in the database.
Table 1 : Average precision for African, horse, dinosaur and flower category for first 10 result images using DCT category African Horse dinosaur flower
Total
Q1 7 10 10 10 relevant images
Q2 8 10 10 9 obtained in
Q3 6 8 10 9
10 retrieved Q4 3 9 10 7 images Q5 7 5 9 8 minimum 3 5 9 7 maximum 8 10 10 10 average 6.2 8.4 9.8 8.6
%precision 62 84 98 86
Table 2 : Average precision for elephant, beach, and food and bus category for first 10 result images using DCT category
Total relevant images obtained in
10 retrieved images minimum maximum average
%precision
Q1
Q2
Q3
Q4
Q5 elephant beach food
8
9
7
9
3
3
9
7.2
72
9
10
8
4
5
4
10
7.2
72
10
7
10
5
2
2
10
6.8
68 bus
9
7
6
5
8
5
9
7
70
Table 3 : Average precision and recall comparison of different categories for first 10 result images using 5 queries of each class
Category
African
Horse
Dinosaur
Flower
Elephant
Beach food bus
Using DCT
Precision
6.2
Recall
.062
8.4
9.8
.084
.098
8.6
7.2
7.2
6.8
7.0
.086
.072
.072
.068
.070

International Journal of Application or Innovation in Engineering & Management (IJAIEM)
Web Site: www.ijaiem.org Email: editor@ijaiem.org, editorijaiem@gmail.com
Volume 1, Issue 2, October 2012
 Average precision including all categories:76.5
ISSN 2319 - 4847
We have presented a new algorithm for digital image search and retrieval in this paper. We have used DCT of each R,
G and B component of images separately to generate feature vectors of dimension (1×3) which is very small in size as compared to using full transform as a feature vector. For all images 3 components i.e. 1 component of R, G and B are considered for feature vector generation. Thus the algorithm is very fast as compared to the algorithms using full transforms. The result of the algorithm is shown in the form of average precision and recall of each class in the Table
1, 2 and 3. Performance of the given system can be increased further using other transforms like FFT, Walsh transform etc.
REFERENCES
[1] Wai-Pak Choi , Kin-Man Lam and Wan-Chi Siu, “An efficient algorithm for the extraction of Euclidean skeleton”
IEEE Transaction on Image processing, 2002 .
[2] Rui Y, Huang T S, Chang S F. “Image retrieval: current techniques, promising directions and open issues”,
Journal of Visual Communication and Image Representation , 1999, 10(I): 39-62
[3] P.S.Suhasini, Dr. K.sri rama krishna, Dr. I. V. Murali krishna,”cbir using color histogram processing” Journal of
Theoretical and Applied Information Technology 2009.
[4] Dr. H.B.Kekre, Dhirendra Mishra, “CBIR using Upper Six FFT Sectors of Color Images for Feature Vector
Generation”, International Journal of Engineering and Technology Vol.2(2), 2010, 49-54
[5] Bai Xue, Liu Wanjun,” Research of Image Retrieval Based on Color “International Forum on Computer Science-
Technology and Applications, 2009.
[ 6]Mustafa Ozden, Ediz Polat,” A color image segmentation approach for content-based image retrieval “Pattern
Recognition Society. Published by Elsevier Ltd 2006
[7] Young Deok Chun; Nam Chul Kim; Ick Hoon Jang;"Content-Based Image Retrieval Using Multiresolution Color and Texture Features," Multimedia, IEEE Transactions on, vol.10, no.6, pp.1073-1084, Oct. 2008.
[8] http://wang.ist.psu.edu/docs/related/Image.orig
[9] Dr. H.B.Kekre, Sudeep D. Thepade, Akshay Maloo, “Image Retrieval using Fractional Coefficients of
Transformed Image using DCT and Walsh Transform”, International Journal of Engineering Science and
Technology , Vol. 2(4), pp.362-371, 2010.
[10] P. Gnanasivam and S. Muttan “Gender Identification Using Fingerprint through Frequency Domain Analysis”
European Journal of Scientific Research, ISSN 1450-216X Vol.59 No.2 (2011), pp.191-199, © Euro Journals
Publishing, Inc. 2011
[11] Nezamabadi-pour, H. and Saryazdi, S., “Object-Based Image Indexing and Retrieval in DCT Domain using
Clustering Techniques”, Transactions on Engineering, Computing and Technology , 2004,
[12 ] J.Z.Wang, J.Li and G.Wiederhold, “SIM PLIcity : semantic sensitive integrated matching for picture libraries”,
IEEE Trans. on Pattern Analysis and Machine Intelligence, 2001,vol.23, no.9, pp.947-963.
AUTHOR
Vibha Bhandari is currently pursuing Masters Degree in Communication System from Chhattisgarh
Swami Vivekananda Technical University Bhilai, India.
Mr. Sandeep B. Patil received his B.E. (E&Tc) and M.E. (Elex) from Dr. B.A.M.U. Aurangabad (M.S).
He is a research scholar and pursuing his Ph.D. in the field of image processing from CSVTU Bhilai. He is working as Associate Professor in the department of Electronics & Telecommunication of Shri
Shankaracharya Group of institutions Bhilai. He has published 16 research papers in various National and
International journals and conferences. His research interest include Biometrics and its applications,
Devnagri character recognition Using neural Network, PCA, HMM, Pattern recognition, Biomedical imaging