Signal & Image Processing : An International Journal (SIPIJ) Vol.3, No.

1, February 2012

Content Based Image Retrieval using Color and Texture
Manimala Singha* and K.Hemachandran$
Dept. of Computer Science, Assam University, Silchar India. Pin code 788011
*n.manimala888@gmail.com,$khchandran@rediffmail.com

ABSTRACT
The increased need of content based image retrieval technique can be found in a number of different domains such as Data Mining, Education, Medical Imaging, Crime Prevention, Weather forecasting, Remote Sensing and Management of Earth Resources. This paper presents the content based image retrieval, using features like texture and color, called WBCHIR (Wavelet Based Color Histogram Image Retrieval).The texture and color features are extracted through wavelet transformation and color histogram and the combination of these features is robust to scaling and translation of objects in an image. The proposed system has demonstrated a promising and faster retrieval method on a WANG image database containing 1000 general-purpose color images. The performance has been evaluated by comparing with the existing systems in the literature.

Keywords
Image Retrieval, Color Histogram, Color Spaces, Quantization, Similarity Matching, Haar Wavelet, Precision and Recall.

1. INTRODUCTION
Research on content-based image retrieval has gained tremendous momentum during the last decade. A lot of research work has been carried out on Image Retrieval by many researchers, expanding in both depth and breadth [1]-[5]. The term Content Based Image Retrieval (CBIR) seems to have originated with the work of Kato [6] for the automatic retrieval of the images from a database, based on the color and shape present. Since then, the term has widely been used to describe the process of retrieving desired images from a large collection of database, on the basis of syntactical image features (color, texture and shape). The techniques, tools and algorithms that are used, originate from the fields, such as statistics, pattern recognition, signal processing, data mining and computer vision. In the past decade, many image retrieval systems have been successfully developed, such as the IBM QBIC System [7], developed at the IBM Almaden Research Center, the VIRAGE System [8], developed by the Virage Incorporation, the Photobook System [9], developed by the MIT Media Lab, the VisualSeek System [10], developed at Columbia University, the WBIIS System [11] developed at Stanford University, and the Blobworld System [12], developed at U.C. Berkeley and SIMPLIcity System [13]. Since simply color, texture and shape features cannot sufficiently represent image semantics, semantic-based image retrieval is still an open problem. CBIR is the most important and effective image retrieval method and widely studied in both academia and industry arena. In this paper we propose an image retrieval system, called Wavelet-Based Color Histogram Image Retrieval (WBCHIR),
DOI : 10.5121/sipij.2012.3104 39

The comparison of experimental result of proposed method with other system [37]-[40] found that. of an image [20]. The color and texture analysis are analyzed by using two level grid frameworks and the shape feature is used by using Gradient Vector Flow. The section 3 describes the color feature extraction. texture and color distribution.The color histogram for color feature and wavelet representation for texture and location information of an image.1. A color histogram is a vector. DBPSP and CHKM to enhance image detection rate and simplify computation of image retrieval. Rao et al. Contourlet Transform and color histogram+contourlet transform. [35] have made a comparative study on image retrieval techniques. The proposed method is given in section 6 and section 7 describes the performance evaluation of the proposed method. The section 4. where each element represents the number of pixels falling in a bin. The features computed on tiles serve as local descriptors of color and texture features. Gabor Transform.3. No. Mainly the texture features of an image are analyzed through statistical. their proposed retrieval system gives better performance than the others. color histogram+gabour transform. and regularity. Although there is no formal definition for texture. [41] proposed CTDCIRS (color-texture and dominant color based image retrieval system). [14] proposed a color-texture and color-histogram based image retrieval system (CTCHIR). color correlogram. February 2012 based on the combination of color and texture features. However the color histogram may fail when the texture feature is dominant in an image [21]. texture and shape features by partitioning the image into tiles. difference between pixels of scan pattern (DBPSP) and color histogram for K-mean (CHKM) respectively and (2) a method for image retrieval by integrating CCM. From the experimental results they found that. [33] and Hung and Dai [34] methods. as color co-occurrence matrix (CCM). Histogram Intersection Distance. They 40 . based on color. They proposed (1) three image features. intuitively this descriptor provides measures of the properties such as smoothness. include Histogram Euclidean Distance. under certain a color space [14-17]. Li and Lee [22] have proposed a ring based fuzzy histogram feature to overcome the limitation of conventional color histogram. The extraction of color features from digital images depends on an understanding of the theory of color and the representation of color in digital images. Color spaces are an important component for relating color to its representation in digital form. Histogram Manhattan Distance and Histogram Quadratic Distance [23-27]. Color is usually represented by color histogram. color coherence vector and color moment. coarseness. The color histogram has been used as one of the feature extraction attributes with the advantage like robustness with respect to geometric changes of the objects in the image. for image retrieval. 2. presents the similarity matching. Finally the experimental work and the conclusions are presented in section 8 and section 9 respectively. they integrated three features like Motif cooccurrence matrix (MCM) and difference between pixels of scan pattern (DBPSP) which describes the texture features and dynamic dominant color (DDC) to extract color feature. Hiremath and Pujari [36] proposed CBIR system based on the color. Texture is also considered as one of the feature extraction attributes by many researchers [28-31].Signal & Image Processing : An International Journal (SIPIJ) Vol. Raghupathi et al. presents the texture feature extraction and the section 5. The transformations between different color spaces and the quantization of color information are primary determinants of a given feature extraction method. a brief review of the related work is presented. structural and spectral methods [32]. RELATED WORK Lin et al. This reduces the processing time for retrieval of an image with more promising representatives. using different feature extraction methods like color histogram. The color histogram feature has been used by many researchers for image retrieval [18 and 19]. The rest of the paper is organized as follows: In section 2. their proposed method outperforms the Jhanwar et al. The distance formula used by many researchers.

3. Color quantization is a process that optimizes the use of distinct colors in an image without affecting the visual properties of an image. the value of each bin gives the number of pixels that has the same corresponding color. [50]. A color histogram for a given image is represented by a vector: Where i is the color bin in the color histogram and H[i] represents the number of pixels of color i in the image. and n is the total number of bins used in color histogram. 4. YUV and HSV [44].…. So far.1 COLOR SPACE SELECTION AND COLOR QUANTIZATION The color of an image is represented. p is the total number of pixels of an image [55]. where each histogram bin corresponds to a color in the quantized color space. without a significant reduction in image quality. No.H 2 . and segmentation for the last four decade. Color is also an intuitive feature and plays an important role in image matching. For a true color image. the distinct number of colors is up to 224 = 16777216 and the direct extraction of color feature from the true color will lead to a large computation.2 Color Histogram A color histogram represents the distribution of colors in an image. Typically.H n .…………H i . Accordingly in the color histogram of an image. 3.……. February 2012 compared their results with the work of Jhanwar et al. COLOR FEATURE The color feature has widely been used in CBIR systems. The extraction of color features from digital images depends on an understanding of the theory of color and the representation of color in digital images. the texture is a powerful low-level feature for image search and retrieval applications.H 3 . through a set of bins.H 1 . each pixel in an image will be assigned to a bin of a color histogram.Signal & Image Processing : An International Journal (SIPIJ) Vol. through any of the popular color spaces like RGB. classification. The normalized color histogram H is given as: Where H = H 0 . In HSV color space the color is presented in terms of three components: Hue (H). It has been reported that the HSV color space gives the best color histogram feature. The color histogram is one of the most commonly used color feature representation in image retrieval. thereby reducing the storage space and enhancing the process speed [53]. the color quantization can be used to represent the image. The original idea to use histogram for retrieval comes from Swain and Ballard [27]. YIQ. still there is a lot of potential for the research. [33] and Hung and Dai [34] and found that their method gives better retrieval results than others. XYZ. among the different color spaces [45]-[49]. The application of the HSV color space in the content-based image retrieval has been reported by Surel et al. In order to reduce the computation.3. who realized the power to identify an object using color is much larger than that of a gray scale. In order to compare images of different sizes.H 2 .H n H i = H = H 0 .1.. 3. there is no unique definition for 41 .H i . color histograms should be normalized. Much work has been done on texture analysis. U*V*W*. TEXTURE FEATURE Like color.H 1 .………. Saturation (S) and Value (V) and the HSV color space is based on cylinder coordinates [51 and 52]. L*a*b*. The effect of color quantization on the performance of image retrieval has been reported by many authors [53 and 54]. because of its easy and fast computation [42]-[43].

≤ t < 1 2 0 . Wavelet transforms extract information from signal at different scales by passing the signal through low pass and high pass filters. 4. the image is decomposed into four frequency sub-bands. Discrete Wavelet Sub-band Decomposition Haar wavelets are widely being used since its invention after by Haar [62]. co-occurrence matrices [59] and Tamura features [60]. The wavelet transform represents a function as a superposition of a family of basis functions called wavelets.0 ≤ t ≤ 1 2 1 .1. because they are the fastest to compute and also have been found to perform well in practice [63]. which applies recursive filtering and sub-sampling. The wavelet transforms can be computed linearly with time and thus allowing for very fast algorithms [28]. otherwise and its scaling function φ(t) can be described as: 42 1 . HL. The common known texture descriptors are Wavelet Transform [57]. February 2012 texture.Signal & Image Processing : An International Journal (SIPIJ) Vol. LH. however. Haar used these functions to give an example of a countable orthonormal system for the space of squareintegrable functions on the real line. because of its better localization and computationally inexpensive properties [30 and 31]. we have used Haar wavelets to compute feature signatures. Wavelets are robust with respect to color intensity shifts and can capture both texture and shape information efficiently. At each level (scale). We have used Wavelet Transform. The Haar wavelet's mother wavelet function ψ(t) can be described as: ψ t = 1 −1 .1 Haar Discrete Wavelet Transforms Discrete wavelet transformation (DWT) [61] is used to transform an image from spatial domain into frequency domain. The wavelet transform computation of a two-dimensional image is also a multi-resolution approach. Gabor-filter [58]. and HH where L denotes low frequency and H denotes high frequency as shown in Figure1. an encapsulating scientific definition as given in [56] can be stated as. Haar wavelets enable us to speed up the wavelet computation phase for thousands of sliding windows of varying sizes in an image. In this paper. Wavelets provide multiresolution capability and good energy compaction. LL. DWT decomposes a signal into a set of Basis Functions and Wavelet Functions. which decomposes an image into orthogonal components. No. Figure 1.3. “Texture is an attribute representing the spatial arrangement of the grey levels of the pixels in a region or image”.

is a precise concept. based on the computation of a similarity function between a pair of images. Step 5.1 Histogram Intersection Distance: Swain and Ballard [27] proposed histogram intersection for color image retrieval. Color quantization is carried out using color histogram by assigning 8 level each to hue. PROPOSED METHOD In this study we are proposing two algorithms for image retrieval based on the color histogram and Wavelet-based Color Histogram. FEATURE SIMILARITY MATCHING The Similarity matching is the process of approximating a solution.0 ≤ t < 1 0 . Calculate the similarity matrix of query image and the image present in the database. Convert RGB color space image into HSV color space. Exactness. Color Histogram Step 1. saturation and value to give a quantized HSV space with 8x8x8=512 histogram bins. and Figure 3. to include the case when the cardinalities of the two histograms are different and expressed as: d min Q i . 6. for the reasons given in [64]. Step 4. The normalized histogram is obtained by dividing with the total number of pixels. D i min |Q i |. we have used the Histogram Intersection Distance method. Repeat step1 to step3 on an image in the database. D i 4 and |Q| and |D| represents the magnitude of the histogram for query image and a representative image in the Database. 5. Step 3. Repeat the steps from 4 to 5 for all the images in the database.1. Retrieve the images.3. No. Intersection of histograms was originally defined as: d = min Q i .1. Step 6.φ t = Signal & Image Processing : An International Journal (SIPIJ) Vol. 43 . and the result is a set of likely values. Step 2. Step 7. In our study. 6. however. by modifying the denominator of the original definition. Many researchers have used different similarity matching techniques [23]-[27]. February 2012 1 . D i |D i | 3 Smith and Chang [55] extended the idea. otherwise 2 5. The block diagrams of the proposed methods are shown in Figure 2.

Green. Green. Step10. 8) Compute the Histogram Similarity computation with distance function Retrieved Images Figure 2. Step12. Retrieve the images. Step4.2. Repeat the steps from 9 to 10 for all the images in the database. No. Combine approximate coefficient of Red. Block diagram of proposed Color Histogram 6. Step8. 8. Color quantization is carried out using color histogram by assigning 8 level each to hue. Assign the weights 0. 0. February 2012 Query Image Image Database Convert RGB into HSV Quantize HSV: (8. and Blue Components from an image. Step5. The normalized histogram is obtained by dividing with the total number of pixels.001 to horizontal and 0. and Blue Component. Convert the approximate. Decompose each Red. 8) Compute the Histogram Convert RGB into HSV Quantize HSV: (8. Calculate the similarity matrix of query image and the image present in the database. Step2.003 to approximate coefficients.3. Extract the Red. Step7. saturation and value to give a quantized HSV space with 8x8x8=512 histogram bins. 44 . Step11.001 to vertical coefficients (experimentally observed values).1. Green. Step9. Green. Blue Component using Haar Wavelet transformation at 1st level to get approximate coefficient and vertical. and Blue Component. horizontal and diagonal detail coefficients. Step3. Wavelet-Based Color Histogram (WBCH). Similarly combine the horizontal and vertical coefficients of Red. Step1.Signal & Image Processing : An International Journal (SIPIJ) Vol. Step6. 8. horizontal and vertical coefficients into HSV plane. Repeat step1 to step8 on an image in the database.

The number of relevant items retrieved is the number of the returned images that are similar to the query image in this case. They are defined as: Precision = = (5) (6) Recall = = Where A represent the number of relevant images that are retrieved. B. 8) Compute the Histogram Similarity matrix computation Retrieved Images Figure 3. PERFORMANCE EVALUATION The performance of retrieval of the system can be measured in terms of its recall and precision. H and V) of RGB into HSV Quantize HSV: (8. February 2012 Query Image Image Decomposition using (Haar Wavelets) Image Database Image Decomposition using (Haar Wavelets) Convert (A. 8.1. Recall measures the ability of the system to retrieve all the models that are relevant. 8) Compute the Histogram Convert (A. (A-approximate coefficient. number of relevant items those were not retrieved. No. the average precision is given by: = /10. The total number of items retrieved is the number of images that are returned by the search engine. 44]. V-vertical detail coefficient). The average precision for the images that belongs to the qth category (Aq) has been computed by [65] = ∈ 7 Where q=1. the number of irrelevant items and the C. 2……10. H and V) of RGB into HSV Quantize HSV: (8. 8 45 . 7. while precision measures the ability of the system to retrieve only the models that are relevant.3.Signal & Image Processing : An International Journal (SIPIJ) Vol. It has been reported that the histogram gives the best performance through recall and precision value [35. Block diagram of proposed Wavelet-Based Color Histogram (WBCH). Finally. 8. H-horizontal detail coefficient.

35.8) color quantization bin and the performance of the proposed image retrieval technique has been evaluated by comparing the results with the results of different authors [14. of each category. Each query returns the top 10 images from database. (8. (4.5 4. (16.4 10 3 8 9 8 7 7 10 6 8 7.8 9 5 6 9 10 8 6 9 6 7 7. (8.3. 46 .4 9 6 7 9 8 8 7 9 5 8 7.6 18. 4) and (18.4 4.2 8. For each query.2. 8. The 10 query retrievals by the proposed method are shown in Figures 5-14. We have followed the image retrieval technique. we examined the precision of the retrieval. Table 1. 3. from the image database. 4).000 images of the Corel stock photo.3 9 5 9 8 8 7 6 10 5 8 7. 3) has been evaluated by randomly selecting 10 query images. The search is usually based on similarity rather than the exact match.8. 8) quantization bin indicates the better retrieval results than the others. 4.6 8. with an average retrieval time as 1min. Precision Using Different Quantization Schemes Category African People Beach Building Buses Dinosaurs Elephants Flowers Horses Mountains Food Average precision 4. The semantic relevance is determined by manual truthing the query image and each of the retrieved images in the retrieval. 8. The average precision (7. and the calculated precision values.8 16.4. 8. 8). The precision values. These results clearly show that the performance of the proposed method is better than the other methods. The quality of the image retrieval. 4). as described in the section 6. EXPERIMENT The proposed method has been implemented using Matlab 7. (8.8.8 9 5 6 9 9 8 6 9 6 9 7.4. with different quantization schemes like (4.Signal & Image Processing : An International Journal (SIPIJ) Vol. 4. as discussed in section 6. February 2012 8.4 9 4 6 9 7 7 7 10 5 8 7. in JPEG format of size 384x256 and 256x386 as shown in Figure 4. has been used to study the image retrieval using (8. calculated by using the equation 5 and also the average precision using the equation 8 are shown in Table 2.3 and tested on a general-purpose WANG database [66] containing 1. 8. 4.8 9 6 6 9 9 9 7 9 6 8 7. 36 and 41] as shown in the Table 2.3.1 on different quantization schemes. using the equation 5.8) value of (8.6 8. No.4 9 6 6 9 8 7 6 9 6 8 7. The effectiveness of the WBCH retrieval method is evaluated by selecting 10 query images under each category of different semantics.8. and average precision using equation 8 are given in the Table 1. 4).8. based on the relevance of the image semantics.4.5 The WBCH method. 8).4.1.

45 0.89 0.61 0.82 0.49 0.92 0.30 0.53 0.65 0.694 0.3.7048 47 .66 0. February 2012 Table 2: Precision of the Retrieval by different methods Classes 1 2 3 4 5 6 7 8 9 10 Category African people Beach Building Buses Dinosaurs Elephants Flowers Horses Mountains Food Average Precision WBCH 0.84 0.73 0.54 0.512 0.62 0.68 0.66 0.578 0.78 0.69 1 0.536 0.91 0.726 [35] 0.95 0.Signal & Image Processing : An International Journal (SIPIJ) Vol.54 0.97 0.72 0.96 0.86 0.704 [36] 0.53 0.6 0.36 0.742 [14] 0.1.899 0.71 0.75 0.984 0.762 CH 0.89 0.43 0.56 0.585 [41] 0.72 0. No.99 0.8 0.61 0.47 0.87 0.93 0.65 0.89 0.77 0.64 0.68 0.75 0.562 0.38 0.76 0.893 0.62 0.93 0.52 0.

No.Signal & Image Processing : An International Journal (SIPIJ) Vol. February 2012 Figure 4.3. Sample of WANG Image Database 48 .1.

February 2012 Figure 5.3. Retrieve results for African People Figure 6. Retrieve results for Beach 49 .1. No.Signal & Image Processing : An International Journal (SIPIJ) Vol.

February 2012 Figure 7.1. Retrieve results for Building Figure 8.Signal & Image Processing : An International Journal (SIPIJ) Vol. No.3. Retrieve results for Bus 50 .

Signal & Image Processing : An International Journal (SIPIJ) Vol. No.1.3. Retrieve results for Elephants 51 . February 2012 Figure 9. Retrieve results for Dinosaurs Figure 10.

Retrieve results for Flowers Figure 12.3. No.1.Signal & Image Processing : An International Journal (SIPIJ) Vol. Retrieve results for Horses 52 . February 2012 Figure 11.

The whole indexing time for the 1000 image database takes 5-6 minutes.3. The experimental result shows that the proposed method outperforms the other retrieval methods in terms of Average Precision. CONCLUSION In this paper. Retrieve results for Mountains Figure 14. Moreover.Signal & Image Processing : An International Journal (SIPIJ) Vol. February 2012 Figure 13. there is a substational increase in the retrieval speed. we presented a novel approach for Content Based Image Retrieval by combining the color and texture features called Wavelet-Based Color Histogram Image Retrieval (WBCHIR).1. As a result. No. Similarity between the images is ascertained by means of a distance function. the computational steps are effectively reduced with the use of Wavelet transformation. Retrieve results for Food 9. 53 .

J.29-40. 2. Belongie. IEEE Workshop on Application of Computer Vision. Malik. Silchar for the award of UGC Fellowship. 70–79. 14. pp-23-32. 1999. pp.F. 12. 16. pp. 2001. G. Pattern Recognition. 15. G. 2007. 9–16. Proceedings of the IEEE Conference. “Database architecture for content-based image retrieval”. 2000. Wiederhold. Li and J. Pattern Anal.The International Society for Optical Engineering. J. Meng.1662. Visual information retrieval. J. 11. REFERENCES 1. Ravi. Vol. and trends of the new age”. pp233–254. Yanker. . M. Worring and S. Kuo. IEEE Transactions on Pattern Analysis and Machine Intelligence. D. “Content-Based Image Retrieval”. 54 5. J. No. R. No. pp. Technical report. pp-262-282. vol. Rui.F. Chan. 9.K. Pass and R. Chen and Y. “Image Indexing Using Color Correlograms” . Wang. 2670. 6. Steele and P.658–665. R Krishna and I. Published in Computer Science Journal. Santini. D. 39-62. “Image Retrieval: Current Techniques. Gupta and R.1. O. Hafne. R. J. 7. IEEE Trans. 1998. Petkovic. 22(12): pp.H. IEEE Transactions on Pattern Analysis and Machine Intelligence. Zang. pp-116-122.W. 1996. “Image retrieval: Ideas. 2009. Jain. “CBIR Using Color Histogram Processing”. Y. In Proceedings of the SPIE . 18. G. Joshi. “Query by Image and Video Content The QBIC System” IEEE Computer. Eakins and M. J. Tapodhir Bhattacharjee (VC). Huang and S. 1995. LI and G. 1992. 20. 21. “Blobworld: image segmentation using expectation-maximization and its application to image querying”. Zabih. JISC Technology Applications Programme. pp. Communications of the ACM 40 (5). Computer Vision and Pattern Recognition. Liu. C. Y. R. Datta. ICCCN’05. W. 2009. pp. “Rotation and Scale Invariant Color Image Retrieval Using Fuzzy Clustering”. We. Scaroff. Pentland. X. 4. 2008. H. Gupta and R. J.112-113. “Content-based Image Indexing and Searching Using Daubechies’ Wavelets”. “VisualSEEk: a fully automated content-based image query system”. 8. Image and Vision Computing vol. A. K. Chinese university of Hong Kong. Firschein and S. A. B. M. 19. P. “Refinement Histogram for Content-Based Image Retrieval”. “Color indexing with weak spatial constraints”.27.1. IS&T/SPIE Conf. pp. J. Promising Directions and Open Issues. SPIE 2670. 3. Flickner. 1997. Vol-40. Proceeding of the 14th International Conference on Computer Communications and Networks. Sixth International Conference on Information Technology: New Generations. Assam University. In SPIE Storage and Retrieval for Image and Video Databases IV.C. 1997. 1349-1380. (4). 6. J. Niblack. D. Smeulders. Li and M. Krishna. Vol. 1996. 8 (8). “Photobook: Content-Based Manipulation for Image Databases”. Chakarvarti and X. Wang.1-60. 1026–1038. 17. Intell. “A Study of Color Histogram Based Image Retrieval”. Gorkani. 1999. S. C. T. Suhasini. M. 1996. “A smart content-based image retrieval system based on color and texture feature”. vol. Q. Jain. Kato. Journal of Theoretical and Applied Information Technology. International Journal of Computer Vision 18 (3). 23.3. Ashley. Huang. 96-102. “Color Distrbution Analysis and Quantization for Image Retrieval”.2. ACM Multimedia. Lee. Huang and S. Lee. pp. influences. R. 1996. “Object Tracking with Dynamic Feature Graph”. Carson. on Storage and Retrieval for Image and Video Databases IV. Picard and S. Wiederhold.Signal & Image Processing : An International Journal (SIPIJ) Vol. M. Tang and H.T. Stricker and A. Dimai. A. 2004.40. pp. 947–963. Journal of Visual Communication and Image Representation. M. Lin. 1996. V. M. Greenspan and J. K. C. 22. 2005. Y. Vol. “A survey of content-based image retrieval with high-level semantics”. Wan and C. 2009. Z. Chang. 10(4): pp. Chang. International Journal on Digital Libraries (IJODL) 1. T. IEEE. 10. S. February 2012 ACKNOWLEDGEMENTS One of us (*) is grateful to the Prof. Lu and W. no. F. D. S. Smith and S. Wang. A. Tae. 9. Jun. R. 311–328. pp. “SIMPLIcity: Semantics-sensitive integrated matching for picture libraries”. Ma. 2002. Graham. “Content Based Image Retrieval at the End of the Early Years”. Dom. Mach. Puerto Rico. ACM computing Survey. 13. H Sawhney. S. D.

Lucena. Banerjee. K. pp-592–598. S. Wenhui and L.L Dewal. J. Prentice Hall. V. 24. 2004. Ardizzoni. Guibas and C. “A Method of defining the similarity of images on the basis of color information”. L. pp-0975-8887. 2010. J. 2011. G. Chen.3. Fuertes. No. W. Technical Report IBM Research. “CTDCIRS: Content based Image Retrieval System based on Dominant Color and Texture Features”. Pattern Recognition Vol. 40. Tomasi. multidimensional scaling. 167–173. Image Processing and Pattern Recognition Vol. G. Y. Management of Data. Huang and S. Second International conference on multimedia and content based image retrieval. Jhanwar.3. H. A. International Journal of Signal Processing. S. Bartolini. G. pp. 36. “Windsurf: Region based Image Retrieval using Wavelets”. Raghupathi. R. 17. P. 22. P. 3. Patella.661-668. IEEE Transactions on Image Processing. pp-65-70. 39. Chan and C. Richard. No. Z. International Journal of Computer Vision. 1998. “Wavelets for texture analysis.R. and A. 2003. 1995. In IWOSS’99. 30. Kumar. 26. Rao. V. and color-based image retrieval”. T. Zhenhua. Ying and Z. and M. pp. X. S. B. “The earth mover’s distance. N. Proceedings of Sixth International Conference on Image Processing and Its Applications. K. Equits. pp-11–32. Kundu and P. 47. Vol. Vol. 7. Ballard. Pattern Recognition.Signal & Image Processing : An International Journal (SIPIJ) Vol. Wouwer and D. “Automated Image Retrieval using Color and Texture”. J. 1252-1267. J. V. N. “WALRUS: A Similarity Retrieval Algorithm for Image Databases”. “Statistical texture characterization from discrete wavelet representation”.147-156. “A Scheme of Color Image Retrieval from Databases”. 2. University of Amsterdam. Pujari. pp-581–585. IEEE. pp. Vol. 2009 Fifth International Conference on Information Assurance and Security.K. 42. Kuo. 27. 28. 38. Bo. In Proceeding. pp-1211–1220. 1997. C. P. 2007. Y. N. Pp-253-257. Chang.22. 1995. IEEE Transactions on Pattern Analysis and Machine Intelligence. Technical Report. H. Wouwer. B. Tokyo Research Laboratory. 44. Dai.9. Y. 1998. Zavidovique. 71. 2. Pp.S.Z. Anand. 1999. “Image retrieval by texture similarity”. 7. Digital Image Processing. V. N. 1999. K. C. 18– No. D. Scheunders and D. Chaudhurib. 37. Rao. “Content based image retrieval using motif co-occurrence matrix”. P. The Journal of Systems and Software. Intelligent Sensory Information Systems. C. Li. No. Vol. “Color and Texture Features for content Based image retrieval”. Govardhan. SPIE 2670. pp665–679. IEEE. “Efficient Color Histogram Indexing for Quadratic Form Distance Functions”. Columbia University. Y. Hiremath and J. Natsev. Gevers. V. G. Z. D. 41. 31. M.323-337. IEEE. D. L Blanca and J. Scheunders. S.9–16. “Image retrieval system based on color-complexity and color-spatial features”. 1997. Rao. Dyck. M. 2001. M. “Content Based Image Retrieval based on Color. No. “IRM: Integrated Region Matching for Image Retrieval”. M. R. 2009. 24. Wan and C. 2004. “An Improving Technique of Color Histogram in Segmentationbased Image Retrieval”. V. the Netherlands. M. A. I. Rubner. 34. pp. 35. M. ICVGIP. No. Vol.6. IEEE Transactions on Pattern Analysis and Machine Intelligence. and M. 1989. P. “IHBM: Integrated Histogram Bin Matching For Similarity Measures of Color Image Retrieval”. R. L. Oct. 2004. Vol. Ioka. Das. 46. 55 25. 2001. 2009. Image and Vision Computing. 45. D. 1991. International Journal of Computer Applications. Swain. 1996. Gonzalez and E. 43. Krishna. J. V. Martinez. July-2123. an overview”. 36. Texture and Shape features using Image and its complement”. “Image Retrieval with Visually Prominent Features using Fuzzy set theoretic Evaluation”. Chen and J. James. HongJiang. P. 32. W. Rastogi and K. “Benchmarking of image feature for content-based retrieval”. S. 33.F. Wiederhold. M. Wang and G. Niblack.W. Smith and S. ACM SIGMOD Int. “A Region-Based Fuzzy Feature Matching Approach to Content Based Image Retrieval”. “Color distribution analysis and quantization for image retrieval”. pp-381-384. Wang. Seetharamanc and B. Color in image Database. pp-395–406.1. 2000. pp. In SPIE Storage and Retrieval for Image and Video Databases IV. Dyck. Conf. February 2012 23. 15th International Conference on Advance Computing and Communications. M. Flickner and W. G. M. L. Proceedings of DARPA Image understanding Workshop. 2002. 1999.W.8. 29. Vol. Shim. In Proceeding of the 8th ACM Intermational Conference on Multimedia. Vol. Livens. Vol. “Color indexing”. Rao and V. . Harpreet.

R. R. IEEE. Vol. Taghizadeh. “Image Retrieval-Based on Color Histogram and performance Evaluation of similarity Measurement”. 1995. 2670. Moghadam. 1998. Pramanik. A. J. 5. Smith and S. Hemachandran. Saadatmand. S. in: IST/SPIE-Storage and Retrieval for Image and Video Databases IV.T. E. 1910. Wu. Shin. “Transform Features for Texture Classification and Discrimination in Large Image Databases. Wan and C. B. CA. L. “Image retrieval with multiresolution color space quantization”.371-378. 55. By Pearson Education. On Systems. R. Gonzalez and R. In Proc. Salesin. 62. 60.edu/docs/related/. 1996. February 2012 48.H. “Comparing the influence of color spaces and metrics in content-based image retrieval”. “Performance analysis of Color Spaces in Image Retrieval”. 2008. 610.. E. “Wavelet correlogram: a new approach for image indexing and retrieval”. J. 2007. “Statistical and structural approaches to texture”. Smith and S. pp. 1990. 52. Assam University Journal of science & Technology. “A new approach to image retrieval with hierarchical color clustering”. “A texture descriptor for browsing and similarity retrieval”. 1994. 1996. A. Gonzalez. E. C. 1996. Digital Image Processing. Yamawaki. 65. No. C. Newsam and H. 2000. Math. IEEE. S. Sural. 58. “Textural features corresponding to visual perception”. 38 pp-2006-2518. 8 Number II 94-104. Chang. pp. pp. 1978. third ed. Annual Conference Series. Finkelstein and D. 1979. Tsang and P. P. Vol. 2002. 8. Prentice Hall. Mori and T. IEEE transactions on circuits and systems for video technology. IEEE International Conference on Image Processing. 786–804. Manjunath. San Jose.33-43. 53. Manimala and K. Chang. Kuo. “Fast Multiresolution image querying”. Annal. Jacobas. pp. no. Vol.4-1990. 50. Vol. Man and Cybern. IEEE. M. Proceedings of the IEEE. Tamura. 54. vol. “Tools and techniques for color image retrieval”. 2011. 64. Eddins. Tsang. S. Woods and S. G. Rouhi and M. 51. 61. 1998. R. H. W. 2011. Wan and C. 8. pp-331-371. 56 . A. C. IEEE. W. C. Vol. pp-277-286.310–349. 59.ist. “Edge gradient method on object color”. Mathias. Vol. TENCON-Digital Signal Processing Application. C. J. IEEE standard glossary of image processing and pattern recognition terminology. X. H. Kuo. Qian and S. 57. E. 66. “Segmentation and Histogram Generation using the HSV Color Space for Image Retrieval”. In Electron Imaging and Multimedia System. Proceeding”. S. 69. T. J. 49. 3. 426-437. IEEE Transactions. Vol. Vol.ICIP. Woods.psu.460-472. A. X.R. Hemachandran... Elsevier Pattern Recognition. Manimala and K. F. Of SiGGRaPH 95. H. R.1. Digital Image Processing Using MALAB. Haralick. 16. 7 Number II 94-104. 56. 2008. S. 63. 67. http://wang. Assam University Journal of science & Technology.Signal & Image Processing : An International Journal (SIPIJ) Vol. Journal of Signal Processing: Image Communication. Zur Theorier der Orthogonalen Funktionensystem.3. pp-407411. H. pp. IEEE Standard. Haar.

and image searching in large databases Prof. K. degrees in Computer Science from Assam University. Silchar in 2005 and 2007 respectively. He obtained his M. since 1998.1.Sc. feature extraction.Sc. His areas of research interest are Image Processing. as a Research Scholar and her area of interest includes image segmentation. Degrees from Indian School of Mines. and M. 57 . Silchar. Assam University. Degree from Sri Venkateswara University.Tech..Signal & Image Processing : An International Journal (SIPIJ) Vol. Manimala Singha received her B. Tirupati and M. of Computer Science.3.Sc.D. February 2012 Authors Ms. Hemachandran is associated with the Dept. No. for her Ph.D. Dhanbad. and Ph. Presently she is working. Software Engineering and Distributed Computing.

Sign up to vote on this title
UsefulNot useful