1.

Introduction
Image Retrieval is the field of study concerned with searching and retrieving digital images from a collection of database. Image retrieval attracts interest among researchers in the fields of image processing, multimedia, digital libraries, remote sensing, astronomy, database applications and others related area. An effective image retrieval system is able to operate on the collection of images to retrieve the relevant images based on the query image which conforms as closely as possible to human perception. Now a days, due to exponential increase in the size of the so called multimedia files in recent years causes retrieval of images from large datasets very crucial.The explosive growth of image data leads to the need of research and development of Image Retrieval. Two major research communities (database management and computer vision) study image retrieval from different perspectives, one being text-based and the other visual based [1].Textbased image retrieval techniques employ text to describe the content of the image while visual based or content-based image retrieval (CBIR) used visual features to describe the content of images. However, Image retrieval researches are moving from keyword, to low level features and to semantic features. From historical perspective, one shall notice that the earlier image retrieval systems are rather text -based search since the images are required to be annotated and indexed accordingly. However, with the substantial increase of the size of images as well as the size of image database, the task of user based annotation becomes very hectic.This motivates the research into content-based image retrieval (CBIR).Content based image retrieval is based on (automated) matching of the features of the query image with that of image database through some image-image similarity evaluation.

2. Background
Image retrieval system is classified into three broad systems as text based image retrieval in which keywords are used as descriptors to retrieve animage. Second is content based image

retrieval in which image features are extracted to form feature vector and then comparison is done. Third is semantic based image retrieval system in which low level

Image Retrieval

Text Based Image Retrieval

Content Based Image Retrieval

Semantic Based Image Retrieval

Fig 1.1 Classification of Image Retrieval System

2.1 Text based Image Retrieval

In conventional image retrieval system, keywords are used as descriptors to index an image however the content of an image is much richer than what any set of keywords can express. Text-based image retrieval techniques employ text to describe the content of the image which often causes ambiguity and inadequacy in performing an image database search and query processing. This problem is due to the difficulty in specifying exact terms and phrases in describing the content of images as the content of an image is much richer than what any set of keywords can express.. In this case there are several difficulties especially redundancy is the main one which is same image that may have different captions and different images may have same keywords. This results in degradation in efficiency of image retrieval. So alternatively, we can efficiently search the images using the content of the image which is incorporated by the CBIR system [3].
Water lilies

Flowers in a pond

Fig 2.1Text Based Image Retrieval

2.2 Content based Image Retrieval

Content-based image retrieval (CBIR) then has been used as an alternative to text based image retrieval. IBM was the first, who take an initiative by proposing query-by image content (QBIC). QBIC developed at the IBM Alma den Research Centre is an open framework and development technology. Unlike keywords-based system, visual features for contents-based system are extracted from the image itself. CBIR can be categorized based on the type of features used for retrieval which could be either low level or high level features. At early years, low level features include colour, texture, shape and spatial relations were used. The CBIR researches were done in retrieving the image on the basis of their visual content as shown in Table 2.1. Table 2.1 Content Based Retrieval Descriptors COLOR Histogram & color moments Region histogram Dominant color Color cluster Bit signature SHAPE Fourier transform Curvature scale space Template matching Convex parts TEXTURE Wavelet transform Edge statistics Gabor filters

Although there are many sophisticated algorithms to describe color, shape and texture features approaches, these algorithms do not satisfied and comfort to human perception. This is mainly due to the unavailability of low level image features in describing high level concepts in the users mind. For an example finding an image of a little girl is playing a ball near the sea. The

only way a machine is able to do isautomatic extraction for the low level features that represented in low level features from images with a good degree of efficiency. Since the low level image features cannot always describe the high level concepts in the users mind such as find an image of a little girl is playing a ball. The only way a machine is able to perform automatic extraction is by extracting the low level features that represented by the color, texture, shape and spatial from images with a good degree of efficiency. The semantic gap which is the lack of correlation between the semantic categories that a user requires and the low-level features that CBIR systems offer has been explored.

Database Image

QueryImag e

Feature Extraction

Similarity Matching

Feature Extraction

Retrieved Images Fig 2.2 Schematic Block Diagram of CBIR

2.3 Semantic based Image Retrieval

Neither a single features nor a combination of multiple visual features could fully capture high level concept of images. Besides, due to the performance of Image retrieval based on low level features are not satisfactory, there is a need for the mainstream of the research converges to retrieval based on semantic meaning by trying to extract the cognitive concept of a human to map the low level image features to high level concept (semantic gap). In addition, representing image content with semantic terms allows users to access images through text query which is more intuitive, easier and preferred by the front end users to express their mind compare with

using images. For example, user queries may be Find an image of sunset rather than find me an image contains red and yellow colours. The semantic extraction and representation of images are shown in Fig 2.3. Semantic representation of images can be done through the process as shown in Fig 2.3. Firstly, the image extraction process will get the low level features of images either by color, shape, textures and spatial. Next, these low level features can be clustered or segmented based on the similar characteristics of the visual features to form some regions representation and next to form objects representation in the images. The regions/objects representation will be annotated with keyword by image annotation process. This annotation process can be done either manually, semi-automatically or automatically. The image then will be represented using semantics and image retrieval can be queried based on high level concept. Semantic content representation has been identified as an important issue to bridge the semantic gap in visual information access. Spatial relationship among these objects/regions can be used to further increase the confidence in image understanding.
H Human Language

Semantic representation Annotated region/objects Region/object representation

Low level features Images

Fig 2.3 Semantic Extractions of Images

3. Features Used for CBIR

3.1 Colour

Color has been the most effective feature and almost all systems employ colours. Although most of the images are in the red, green, blue (RGB) color space, this space is only rarely used

for indexing and querying as it does not correspond well to the human color perception. It only seems reasonable to be used for images taken under exactly the same conditions each time such as trademark images. Other spaces such as hue, saturation, value (HSV)or the CIE Lab and Luv spaces are much better with respect to human perception and are more frequently used. This means that differences in the color space are similar to the differences between colors that humans perceive. Much effort has also been spent on creating color spaces that are optimal with respect to lighting conditionsor that are invariant to shades and other influences such as viewing position. This allows identifying colors even under varying conditions but on the other hand information about the absolute colors is lost. 3.1.1 HSV histogram In the HSV colour space[2] , hue is used to distinguish colours, saturation is the percentage of white light added to a pure colour and value refers to perceived light intensity. In the HSV color space each component occupies a large range of values. We divided Hue into eight parts Saturation and intensity is divided into three parts seprately in accordance with the human eyes to distinguish. Color histogram is derived by first quantizing colors in the image into 72 bin in HSV color space. Then counting the number of image pixels in each bin.

Fig 1.5: HSV Color Space

3.2 Texture

Texture is that innate property of all surfaces that describes visual patterns, each having properties of homogeneity. It contains important information about the structural arrangement of the surface, such as; clouds, leaves, bricks, fabric, etc. It also describes the relationship of the surface to the surrounding environment [2]. In short, it is a feature that describes the distinctive physical composition of a surface. Texture feature is a kind of visual characteristics that does not rely on color or intensity and reflects the intrinsic phenomenon of images. It is the total of all the intrinsic surface properties. Many objects in an image can be distinguished solely by their texture without any other information. In our proposed method we have used the statistic texture feature using gray-level co-occurrence matrix(GLCM). GLCM is created in four directions with the distance between pixels as one. Texture features are extracted from the statistics of this matrix. GLCM is composed of the probability value, it is defined by which expresses the probability of the couple pixels at Texture properties include Energy Contrast Correlation Entropy Roughness

Figure 3.1: Examples of Textures

There are three principal approaches used to describe texture; statistical, structural and spectral Statistical techniques characterize textures using the statistical properties of the grey levels of the points/pixels comprising a surface image. Typically, these properties are computed using: the grey level co-occurrence matrix of the surface, or the wavelet transformation of the surface. Structural techniques characterize textures as being composed of simple primitive structures called texels (or texture elements). These are arranged regularly on a surface according to some surface arrangement rules. Spectral techniques are based on properties of the Fourier spectrum and describe global periodicity of the grey levels of a surface by identifying high-energy peaks in the Fourier spectrum [9]. For optimum classification purposes, what concern us are the statistical techniques of characterization This is because it is these techniques that result in computing texture properties The most popular statistical representations of texture are: 3.2.1 Gray Level Co-occurrence Matrix Originally proposed by R.M. Haralick, the co-occurrence matrix representation of texture features explores the grey level spatial dependence of texture [2]. A mathematical definition of the co-occurrence matrix is as follows [4]: Given a position operator P(i,j), let A be an n x n matrix

whose element A[i][j] is the number of times that points with grey level (intensity) g[i] occur, in the position specified by P, relative to points with grey level g[j].

Let C be the n x n matrix that is produced by dividing A with the total number of point pairs that satisfy P. C[i][j] is a measure of the joint probability that a pair of points satisfying P will have values g[i], g[j].

C is called a co-occurrence matrix defined by P.

Examples for the operator P are: i above j, or i one position to the right and two below j, etc. [4]

This can also be illustrated as follows Let t be a translation, then a co-occurrence matrix Ct of a region is defined for every grey-level (a, b) by [1]:

Ct (a , b) card {( s, s t ) R 2 | A[ s] a , A[ s t ] b}
Here, Ct(a, b) is the number of site-couples, denoted by (s, s + t) that are separated by a translation vector t, with a being the grey-level of s, and b being the grey-level of s + t. For example; with an 8 grey-level image representation and a vector t that considers only one neighbour, we would find .

Figure 3.2: Image example

Fig 3.3: Classical Co-occurrence matrix

At first the co-occurrence matrix is constructed, based on the orientation and distance between image pixels. Then meaningful statistics are extracted from the matrix as the texture representation. Haralick proposed the following texture features : 1. Angular Second Moment 2. Contrast 3. Correlation 4. Variance 5. Inverse Second Differential Moment 6. Sum Average 7. Sum Variance 8. Sum Entropy 9. Entropy 10. Difference Variance 11. Difference Entropy 12. Measure of Correlation 1

13. Measure of Correlation 2 14. Local Mean Hence, for each Haralick texture feature, we obtain a co-occurrence matrix. These cooccurrence matrices represent the spatial distribution and the dependence of the grey levels within a local area. Each (i,j) th entry in the matrices, represents the probability of going from one pixel with a grey level of 'i' to another with a grey level of 'j' under a predefined distance and angle. From these matrices, sets of statistical measures are computed, called feature vectors [11].

3.3 Image Sub-block

Image sub-block mean partitioned the image into blocks. In many technique image is partitioned into 6(2*3) equal sized sub-blocks. The size of sub-block in an image of size 256*384 is 128*128. The images with other than 256*384 sizes are resized to 256*384. Using HSV colour space we can extract the color features of each sub-block according to hue, saturation and value. we can use GLCM (Gray-level co-occurrence Matrix) for texture feature extraction of each block . And then construct a combined feature vector for colour and texture. Using Euclidean distance we can find the distance between query image and target image and retrieve the images from database.

4. Application Areas
A wide range of possible applications for CBIR technology has been identified. Potentially fruitful areas include:

5. Conclusions
This report reviewed the main components of a content based image retrieval system, including image feature representation, indexing, query processing, and query-image matching and user's interaction. It has been acknowledged that HSV colour histogram is one the technique for colour feature extraction and GLCM is one of the technique that can be used for texture extraction of an image sub-block. This technique give better result because here we use combined feature vector (Colour, texture) of an image sub-block.

References

[1] Kavitha, Ch, B. Prabhakara Rao, and A. Govardhan, "Image Retrieval based on combined features of image sub-blocks", International Journal on Computer Science and Engineering 3.4: 1429-1438, 2011. [2] Mustaffa, Mas Rina, Fatimah Ahmad, Rahmita Wirza OK Rahmat, and Ramlan Mahmod, "Contentbased image retrieval based on color-spatial features" ,Malaysian Journal of Computer Science 21, no. 1 (2008): 1-12. [3] Vimina, E. R., and K. Poulose Jacob, "A Sub-block Based Image Retrieval Using Modified Integrated Region Matching." (2013). [4] Lin, T. X., and Chung-Shen Hung, "Quadrant motif approach for image retrieval" In Proceedings of the 14th International Conference on Central Europe on Computer Graphics, Visualization and Computer Vision, Plzen, Czech Republic, pp. 209-215. 2006. [5] Pujari, Jagadeesh, and P. Hiremath, "Content Based Image Retrieval based on Color, Texture and Shape features using Image and its complement, " International Journal of Computer Science and Security [6] Kavitha, Ch, Dr B. Prabhakara Rao, and Dr A. Govardhan, "Image retrieval based on color and texture features of the image sub-blocks," International Journal of Computer Applications 15, no. 7 (2011). [7] Ying Liua, Dengsheng Zhanga, Guojun Lua, Wei-Ying Mab, A survey of content-based image retrieval with high-level semantics, Pattern Recognition Society,Published by Elsevier Ltd.2006. [8] T. Dharani, Department of Computer Science Periyar University et al., A Survey on Content Based Image Retrieval,International Conference on Pattern Recognition, Informatics and Mobile Engineering, IEEE, Pages 485-490, February 2012.

Contents 1. Introduction ...................................................................................... 1 2. Background ....................................................................................... 1 2.1 Text based Image Retrieval .......................................................... 2 2.2 Content based Image Retrieval .................................................... 3 2.3 Semantic based Image Retrieval .................................................. 4 3. Features Used for CBIR ..................................................................... 5 3.1 Colour ........................................................................................... 5 3.1.1 HSV histogram ........................................................................... 6 3.2 Texture ......................................................................................... 7 4. Application Areas ............................................................................ 11 5. Conclusions ..................................................................................... 12 References.13