WAVELET BASED COLOR HISTOGRAM IMAGE...
Transcript of WAVELET BASED COLOR HISTOGRAM IMAGE...
International Journal of Science, Engineering and Technology Research (IJSETR), Volume 3, Issue 5, May 2014
ISSN: 2278 – 7798 All Rights Reserved © 2014 IJSETR 1524
Abstract— In various application domains such as
education, crime prevention, commerce, and biomedicine, the
volume of digital data is increasing rapidly. The problem
appears when retrieving the information from the storage media.
Content-based image retrieval systems aim to retrieve images
from large image databases similar to the query image based on
the similarity between image features. We present a CBIR system
that uses the color feature as a visual feature to represent the
images. We use the images from the WANG database that is
widely used for CBIR performance evaluation. The database
contains color images, so we use the RGB color space to
represent the images. So we make use of 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.
Index Terms—Image Retrieval, Color Histogram, Color Spaces,
Quantization, Similarity Matching, Haar Wavelet,
Precision and Recall
I. INTRODUCTION
With the development of the Internet, and the availability of
image capturing devices such as digital cameras, image
scanners, the size of digital image collection is increasing
rapidly. Efficient image searching, browsing and retrieval
tools are required by users from various domains, including
remote sensing, fashion, crime prevention, publishing, medicine, architecture, etc. For this purpose, many general
purpose image retrieval systems have been developed.
There are two frameworks: text-based and content-based.
The text-based approach can be traced back to 1970s. In
such systems, the images are manually annotated by text
descriptors, which are then used by a database management
system. (DBMS) to perform image retrieval.manual
annotation. The second is the inaccuracy in annotation due
to the subjectivity of human perception. To overcome these
disadvantages in text-based retrieval system, content-based
image retrieval (CBIR) was introduced in the early 1980s. In CBIR, images are indexed by their visual content, such
as color, texture, shapes. The fundamental difference
`
between content-based and text-based retrieval systems is
that the human interaction is an essential part of the latter
system. As a result, content-based image retrieval (CBIR)
from unannotated image databases has been a fast growing
research area recently.
The term Content-based image retrieval was originated in
1992, when it was used by T. Kato to describe experiments
into automatic retrieval of images from a database, based on
the colors and texture present. Since then, this term has
been used to describe the process of retrieving desired
images from a large collection on the basis of syntactical
image features. The techniques, tools and algorithms that
are used originate from fields such as statistics, pattern
recognition, signal processing, and computer vision.
1.1 Architecture Of CBIR
Content-based Image Retrieval (CBIR) is the searching of
an image database based on what is captured by the
individual images of the collection. There are various ways of implementing these searches and they will be explored
shortly. Content-based image retrieval (CBIR), the image
databases are indexed with descriptors derived from the
visual content of the images [2]. Most of the CBIR systems
are concerned with approximate queries where the aim is to
find images visually similar to a specified target image.
In most cases the aim of CBIR systems is to replicate
human perception of image similarity as well as possible.
The process of CBIR consists of the following stages:
(1) Image acquisition: to acquire a digital image.
Image Database: It consists of the collection of n
number of images depends on the user range and
choice.
(2) Image preprocessing: To improve the image in ways that increases the chances for success of the other
processes. The image is first processed in order to extract
the features, which describe its contents. The processing
involves filtering, normalization, segmentation, and object
identification. Like, image segmentation is the process of
dividing an image into multiple parts. The output of this
stage is a set of significant regions and objects.
WAVELET BASED COLOR HISTOGRAM
IMAGE RETRIEVAL
UdayaTheja.V, Sangamesh, Dr.Rajshekhar Ghogge
International Journal of Science, Engineering and Technology Research (IJSETR), Volume 3, Issue 5, May 2014
ISSN: 2278 – 7798 All Rights Reserved © 2014 IJSETR 1525
(3) Feature Extraction: Features such as shape, texture,
color, etc. are used to describe the content of the image. The
features further can be classified as low-level and high-level
features. In this step visual information is extracts from the
image and saves them as features vectors in a features
database .For each pixel, the image description is found in the form of feature value (or a set of value called a feature
vector) by using the feature extraction .These feature
vectors are used to compare the query with the other images
and retrieval.
(4) Similarity Matching: The information about each image
is stored in its feature vectors for computation process and
these feature vectors are matched with the feature vectors of
query image (the image to be search in the image database
whether the same image is present or not or how many are
similar kind images are exist or not) which helps in
measuring the similarity. This step involves the matching of the above stated features to yield a result that is visually
similar with the use of similarity measure method called as
Distance method. Here is different distances method
available such as Euclidean distance, City Block Distance,
Canberra Distance.
(5) Resultant Retrieved images: It searches the previously
maintained information to find the matched images from
database. The output will be the similar images having
same or very closest features as that of the query image.
(6) User interface and feedback which governs the display
of the outcomes, their ranking, the type of user interaction
with possibility of refining the search through some
automatic or manual preferences scheme etc.
Fig -1.1: CBIR System and its various components
A typical CBIR uses the contents of an image to represent
and access. CBIR systems extract features (color, texture, and shape) from images in the database based on the value
of the image pixels. These features are smaller than the
image size and stored in a database called feature database.
Thus the feature database contains an abstraction (compact
form) of the images in the image database; each image is
represented by a compact representation of its contents
(color, texture, shape and spatial information) in the form of
a fixed length real-valued multi-component feature vectors
or signature. This is called off-line feature extraction.
When the user submits a query image to the CBIR system,
the system automatically extracts the features of the query
image in the same way as it does for the image database.
The distance (similarity) between the feature vector of the
query image and the feature vectors stored in the feature
database are computed. The system will sort and retrieve
the best similar images according to their similarity values.
This is called on-line image retrieval. The main advantage of using CBIR system is that the system uses image features
instead of using the image itself. So, CBIR is cheap, fast
and efficient over image search methods.
We propose an image retrieval system, called Wavelet-Based Color Histogram Image Retrieval (WBCHIR), based
on the combination of color and texture features. The color
histogram for color feature and wavelet representation for
texture and location information of an image. This reduces
the processing time for retrieval of an image with more
promising representatives.
1.2 The Importance Of Content Based Image Retrieval
To search for an image, we use some text or a keyword that
describes the image to retrieve it. This method is not good
for image retrieving, because in that case, every image must
have a powerful complete description and then must match
the words we use to search. Unfortunately, we have huge
image databases, and it is illogical to describe every image
in the database with a good complete description and when we retrieve the images, the system will often miss some
images and will retrieve images that don't relate to what we
need. From this point, we want to find a new technique and
use it to retrieve images depending on its content not its
description.
To solve the problem of searching for an image using text,
we will use the content of the image to search and retrieve
it. CBIR is a technique to search and retrieve images. A
content-based retrieval system processes the information
contained in image data and creates an abstraction of its
content in terms of visual attributes. These attributes are
color, shape, and texture. Any query operations deal with
this abstraction rather than with the image itself. Thus,
every image inserted into the database is analyzed, and a
compact representation of its content is stored in a feature
vector, or signature.
To retrieve an image, the query image must compare with
other images in the database for similarity. Similarity
comparison uses the image representation. Representation of an image includes extracting some features. Features
extracted from an image can be color, texture, or shape. The
similarity is to calculate the difference between the images'
features. For this point, CBIR has several advantages
comparing with other approaches such we have mentioned
such as text-based retrieval.
International Journal of Science, Engineering and Technology Research (IJSETR), Volume 3, Issue 5, May 2014
ISSN: 2278 – 7798 All Rights Reserved © 2014 IJSETR 1526
1.3 Applications
CBIR concepts have been used widely in many real
applications. Most of our fields need image processing and
retrieving such as medical, architectural, criminal, and in
the web. For the medical field, CBIR is used for diagnosis by identifying similar past cases. In critical buildings, this
technique is used for finger print or retina scanning for
privileges. The most important application that uses CBIR
is the web. Many web applications provide searching and
retrieving images based on their contents. In general,
retrieving images based on their content becomes serious
and important techniques in most of the human
applications. Potentially fruitful areas include:
Crime Prevention
Nowadays, police forces keep large archives of evidence for past suspects, including facial photographs and
fingerprints. When a crime is happened, they take the
evidence from the scene of the crime and compare it with
the records in their archives. They use CBIR systems to get
their results. The most import thing when designing these
systems is the ability to search an entire database to find the
closest matching records instead of matching against only a
single stored record.
Medical Diagnosis
Modern medicine depends on diagnostic methods
such as radiology, histopathology, and computerized
tomography. These diagnostic methods have resulted in a
large number of important medical images that most
hospitals stored. Now, there is a great interest to use of
CBIR methods to aid diagnosis by identifying similar past
cases.
Home Entertainment
Most home entertainment is images such as
holiday images, festivals and videos such as favorite programs and movies. CBIR methods can be used for image
management. Now, number of large organizations devotes
large development effort to design simple software for
retrieval with affordable price.
Web Searching
One of the most CBIR applications is the web
searching. Uses face problems when they search for certain
images by some image description. Many search engines
use the text-based search and for many times the results are not stratified by the user. Some content-based search
engines are developed so that the user submits the query
image and the search engine retrieve the most similar
images. Some content-based search engines provide a
relevance feedback facility to refine search results.
Numerous commercial and experimental CBIR systems are
now available, and many web search engines are now
equipped with CBIR facilities, as for example Alta Vista,
Yahoo and Google.
II. FUNDAMENTALS OF IMAGE RETRIEVAL
The main idea behind CBIR systems is to allow users to
find images that are visually similar to the query image.
Similar may have different meanings. Some users may be
interested in some image regions. Others are interested in some shapes and the color of them. Therefore, different
needs mean different methods for similarity. To allow
different methods for similarity, different image descriptors
are needed. Image descriptors may account for different
properties of images. Image descriptors mean image
features. A feature means anything that is localized,
meaningful and detectable. If we talk about image features,
we mean objects in that image such as corners, lines,
shapes, textures, and motions. Features extracted from an
image describe and define the content of that image.
Intuitively, the most direct method to compare two images
is to compare the pixels in one image to the corresponding
pixels in the other image. Clearly, this method is not
feasible, because images may have different size that
applications cannot determine which pixels from one image correspond to which pixels in the other image. Another
reason is the computational complexity [4]. When a system
wants to match two images by comparing pixel by pixel, it
will take a long time. This is just for two images.
Nowadays, we talk about thousands of images stored in
databases that are used for image retrieving. Comparing
images using their pixels is time consuming. More powerful
method is to use image features instead of using the original
pixel values because of the significant simplification of
image representation, and the easy way to compare images
using their features.
A wide variety of features had been considered for image
retrieval. Color, texture, and shape are some image features
that can be used to describe an image. However, no
particular feature is most suitable for retrieving all types of
images [5]. Color images need color features that are most suitable to describe them. Images containing visual
patterns, surface properties, and scene need texture features
to describe them. In reality, no one particular feature can
describe an image completely. Many images have to be
described by more than one feature. For example, color and
texture features are best features to describe natural scenes.
Features extracted from the image are used for computing
the similarity between images. Some measurement methods
are used to calculate the similarity between images. In this
chapter, we will define image features, explaining their
properties. We introduce some methods for similarity
measures.
2.1 Feature Extraction
Feature extraction means obtaining useful information that
can describe the image with its content. We mean by image
features the characteristic properties. For example, the
image of a forest can be described by its green color and
International Journal of Science, Engineering and Technology Research (IJSETR), Volume 3, Issue 5, May 2014
ISSN: 2278 – 7798 All Rights Reserved © 2014 IJSETR 1527
some texture of trees. Objects in the image can be
considered as shapes that can be a feature for the image. To
describe an image, we have to consider its main features.
Selecting image features is an important step so that it can
represent the content of the image very well. Color and
texture are some features considered for content image
description.
2.2 Color
Color is the sensation caused by the light as it interacts with our eyes and brain. Color features are the fundamental
characteristics of the content of images. Human eyes are
sensitive to colors, and color features enable human to
distinguish between objects in the images. Colors are used
in image processing because they provide powerful
descriptors that can be used to identify and extract objects
from a scene. Color features provide sometimes powerful
information about images, and they are very useful for
image retrieval.
To facilitate the specification of colors in some standard,
color spaces (also called color models or color systems) are
proposed. A color space is a specification of a coordinate
system and a subspace within the system where each color
is represented by a single point. Today, most color spaces in
use are oriented toward hardware (such as for color monitors and printers) or toward software for applications
where color manipulation is the target. In most digital
image processing, RGB (red, green, blue) color space is
used in practice for color monitors and CMY (cyan,
magenta, yellow) color space is used for color printing. In
our work, we are focusing on the RGB color space.
To extract the color features from the content of an image,
we need to select a color space and use its properties in the
extraction. In common, colors are defined in three
dimensional color space. The purpose of the color space is
to facilitate the specification of colors in some standard,
accepted way. Several color spaces are used to represent
images for different purposes. The RGB color space is the
most widely used color space. RGB stands for Red, Green,
and Blue. RGB color space combines the three colors in different ratio to create other colors. In digital image
purposes, RGB color space is the most prevalent choice.
The main drawback of the RGB color space is that it is
perceptually non uniform. We can imagine the RGB color
space as a unit cube with red, green, and blue axes. Any
color in the RGB color space can be represented by a vector
of three coordinates. To overcome the drawback of the
RGB color space, different color spaces are proposed.
The HSx color space is commonly used in digital image
processing that converts the color space of the image from
RGB color space to one of the HSx color spaces. HSx color
space contains the HSI, HSV, HSB color spaces. They are
common to human color perception. HS stands for Hue and
Saturation. I, V, and B stand for Intensity, Value, and
Brightness, respectively. The different difference between
them is their transformation method from the RGB color
space. Hue describes the actual wavelength of the color.
Saturation is the measure of the purity of the color. For
example, red is 100% saturated color, but pink is not 100%
saturated color because it contains an amount of white.
Intensity describes the lightness of the color. HSV color space is the most widely used when converting the color
space from RGB color space.
2.2.1 Methods of Representation
The main method of representing color information of
images in CBIR systems is through color histograms. A
color histogram is a type of bar graph, where each bar
represents a particular color of the color space being used.
In Mat Lab for example you can get a color histogram of an
image in the RGB or HSV color space. The bars in a color
histogram are referred to as bins and they represent the x-
axis. The number of bins depends on the number of colors
there are in an image. The y-axis denotes the number of
pixels there are in each bin. In other words how many
pixels in an image are of a particular color.
An example of a color histogram in the HSV color space
can be seen with the following image:
Fig -2.1: Sample Image and its Corresponding HistogramTo view a
histogram numerically one has to look at the color map or the numeric
representation of each bin.
International Journal of Science, Engineering and Technology Research (IJSETR), Volume 3, Issue 5, May 2014
ISSN: 2278 – 7798 All Rights Reserved © 2014 IJSETR 1528
Table -2.1: Color Map and Number of pixels for the Previous Image.
As one can see from the color map each row represents the
color of a bin. The row is composed of the three coordinates
of the color space. The first coordinate represents hue, the
second saturation, and the third, value, thereby giving HSV.
The percentages of each of these coordinates are what make
up the color of a bin. Also one can see the corresponding
pixel numbers for each bin, which are denoted by the blue
lines in the histogram.
Quantization in terms of color histograms refers to the
process of reducing the number of bins by taking colors that
are very similar to each other and putting them in the same
bin. By default the maximum number of bins one can obtain using the histogram function in MatLab is 256. For
the purpose of saving time when trying to compare color
histograms, one can quantize the number of bins [3].
Obviously quantization reduces the information regarding
the content of images but as was mentioned this is the
tradeoff when one wants to reduce processing time.
There are two types of color histograms, Global color
histograms (GCHs) and Local color histograms (LCHs). A
GCH represents one whole image with a single color
histogram. An LCH divides an image into fixed blocks and
takes the color histogram of each of those blocks. LCHs
contain more information about an image but are
computationally expensive when comparing images. “The
GCH is the traditional method for color based image
retrieval. However, it does not include information concerning the color distribution of the regions of an image.
Thus when comparing GCHs one might not always get a
proper result in terms of similarity of images.
2.3 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. In short, it is a feature that describes the distinctive physical composition of a surface.
Texture properties include: Coarseness, Contrast,
Directionality, Line-likeness, Regularity, and Roughness.
Texture is one of the most important defining features of an image. It is characterized by the spatial distribution of gray
levels in a neighborhood. In order to capture the spatial
dependence of gray-level values, which contribute to the
perception of texture, a two-dimensional dependence
texture analysis matrix is taken into consideration. This
two-dimensional matrix is obtained by decoding the image
file; jpeg, bmp, etc.
2.3.1 Methods of Representation
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.
Fig -2.2: Examples of Textures
Clouds
Bricks
Rocks
International Journal of Science, Engineering and Technology Research (IJSETR), Volume 3, Issue 5, May 2014
ISSN: 2278 – 7798 All Rights Reserved © 2014 IJSETR 1529
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. 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: Co-
occurrence Matrix, Tamura Texture, and Wavelet
Transform.
III. IMAGE RETRIEVAL BASED ON CONTENT
We introduce our proposed CBIR system. In our proposed
system, we will extract some color features to represent the
image and use these features to compare between the
images.
3.1 Color Feature Extraction
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 color
histogram is one of the most commonly used color feature
representation in image retrieval. The power to identify an
object using color is much larger than that of a gray scale.
3.1.1 Color Space Selection And Color Quantization
The color of an image is represented, through any of the
popular color spaces like RGB, XYZ, YIQ, L*a*b*,
U*V*W*, YUV and HSV. It has been reported that the
HSV color space gives the best color histogram feature,
among the different color spaces. In HSV color space the
color is presented in terms of three components: Hue (H),
Saturation (S) and Value (V) and the HSV color space is
based on cylinder coordinates.
Color quantization is a process that optimizes the use of
distinct colors in an image without affecting the visual
properties of an image. 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. In order to reduce the computation, the
color quantization can be used to represent the image,
without a significant reduction in image quality, thereby
reducing the storage space and enhancing the process
speed.
3.1.2 Color Histogram
A color histogram represents the distribution of
colors in an image, through a set of bins, where each
histogram bin corresponds to a color in the quantized color
space. A color histogram for a given image is represented
by a vector:
H = {H[0] , H[1] , H[2] , H[3] ,
H[4],……….H[i],…….,H[n]}.
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.
Typically, each pixel in an image will be assigned to a bin
of a color histogram. Accordingly in the color histogram of an image, the value of each bin gives the number of pixels
that has the same corresponding color. In order to compare
images of different sizes, color histograms should be
normalized. The normalized color histogram H` is given as:
Where p is the total number of pixels of an image.
3.2 Texture Feature Extraction
Like color, the texture is a powerful low-level feature for
image search and retrieval applications. Much work has
been done on texture analysis, classification, and
segmentation for the last four decade, still there is a lot of
potential for the research. So far, there is no unique
definition for texture; however, an encapsulating scientific
definition as given in can be stated as, “Texture is an attribute representing the spatial arrangement of the grey
levels of the pixels in a region or image”. The common
known texture descriptors are Wavelet Transform, Gabor-
filter, co-occurrence matrices and Tamura features. We
have used Wavelet Transform, which decomposes an image
into orthogonal components, because of its better
localization and computationally inexpensive properties.
3.2.1 Haar Discrete Wavelet Transforms
Discrete wavelet transformation (DWT) is used to
transform an image from spatial domain into frequency
domain. The wavelet transform represents a function as a
superposition of a family of basis functions called wavelets.
Wavelet transforms extract information from signal at
different scales by passing the signal through low pass and high pass filters. Wavelets provide multi-resolution
capability and good energy compaction. Wavelets are
robust with respect to color intensity shifts and can capture
both texture and shape information efficiently. The wavelet
transforms can be computed linearly with time and thus
allowing for very fast algorithms. DWT decomposes a
signal into a set of Basis Functions and Wavelet Functions.
The wavelet transform computation of a two-dimensional
image is also a multi-resolution approach, which applies
recursive filtering and sub-sampling. At each level (scale),
the image is decomposed into four frequency sub-bands, LL, LH, HL, and HH where L denotes low frequency and H
denotes high frequency as shown in Figure.
International Journal of Science, Engineering and Technology Research (IJSETR), Volume 3, Issue 5, May 2014
ISSN: 2278 – 7798 All Rights Reserved © 2014 IJSETR 1530
Fig -3.1: Discrete Wavelet Sub-band
Decomposition
Haar wavelets are widely being used since its invention
after by Haar. Haar used these functions to give an example
of a countable orthonormal system for the space of square-
integrable functions on the real line. Here, we have used
Haar wavelets to compute feature signatures, because they
are the fastest to compute and also have been found to
perform well in practice. Haar wavelets enable us to speed
up the wavelet computation phase for thousands of sliding
windows of varying sizes in an image. The Haar wavelet's
mother wavelet function can be described as:
And its scaling function can be described as:
3.3 Feature Similarity Matching
The Similarity matching is the process of approximating a
solution, based on the computation of a similarity function
between a pair of images, and the result is a set of likely
values. Exactness, however, is a precise concept.
3.3.1 Histogram Intersection Distance
Swain and Ballard [4] proposed histogram intersection for
color image retrieval. Intersection of histograms was
originally defined as:
Smith and Chang [6] extended the idea, by modifying the
denominator of the original definition, to include the case
when the cardinalities of the two histograms are different
and expressed as:
and |Q| and |D| represents the magnitude of the histogram
for query image and a representative image in the Database.
IV. PROPOSED METHODOLOGY
Proposing two algorithms for image retrieval based on the
color histogram and Wavelet-based Color Histogram.
4.1 Color Histogram
Input: Query Image.
Output: Retrieved Images
Method: 1. Convert RGB color space image into HSV color space.
2. Color quantization is carried out using color histogram
by assigning 8 level each to
hue, saturation and value to give a quantized HSV space
with 8x8x8=512 histogram
bins.
3. The normalized histogram is obtained by dividing with
the total number of pixels.
4. Repeat step1 to step3 on an image in the database.
5. Calculate the similarity matrix of query image and the
image present in the database. 6. Repeat the steps from 4 to 5 for all the images in the
database.
7. Retrieve the images.
Fig -4.1: Block diagram of proposed Color Histogram
4.2 Wavelet-Based Color Histogram (Wbch)
Input: Query Image.
Output: Retrieved Images.
Method:
1. Extract the Red, Green, and Blue Components from an
image.
International Journal of Science, Engineering and Technology Research (IJSETR), Volume 3, Issue 5, May 2014
ISSN: 2278 – 7798 All Rights Reserved © 2014 IJSETR 1531
2. Decompose each Red, Green, Blue Component using
Haar Wavelet transformation at
1st level to get approximate coefficient and vertical,
horizontal and diagonal detail
Coefficients.
3. Combine approximate coefficient of Red, Green, and Blue Component.
4. Similarly combine the horizontal and vertical coefficients
of Red, Green, and Blue
Component.
5. Assign the weights 0.003 to approximate coefficients,
0.001 to horizontal and 0.001 to
Vertical coefficients (experimentally observed values).
6. Convert the approximate, horizontal and vertical
coefficients into HSV plane.
7. Color quantization is carried out using color histogram
by assigning 8 level each to hue,
Saturation and value to give a quantized HSV space with 8x8x8=512 histogram bins.
8. The normalized histogram is obtained by dividing with
the total number of pixels.
9. Repeat step1 to step8 on an image in the database.
10. Calculate the similarity matrix of query image and the
image present in the database.
11. Repeat the steps from 9 to 10 for all the images in the
database.
12. Retrieve the images.
Fig -4.2: Block diagram of proposed Wavelet-Based Color
Histogram (WBCH). (A-approximate
coefficient, H-horizontal detail coefficient, V-vertical detail
coefficient).
4.3 Performance Evaluation
The performance of retrieval of the system can be measured
in terms of its recall and precision. Recall measures the
ability of the system to retrieve all the models that are
relevant, while precision measures the ability of the system
to retrieve only the models that are relevant. It has been
reported that the histogram gives the best performance
through recall and precision value They are defined as:
Where A represent the number of relevant images that are
retrieved, B, the number of irrelevant items and the C,
number of relevant items those were not retrieved. The number of relevant items retrieved is the number of the
returned images that are similar to the query image in this
case. The total number of items retrieved is the number of
images that are returned by the search engine.
The average precision for the images that belongs to the qth
category (Aq) has been computed by
4.4 Example
Fig -4.3: Retrieval Results of Bus
International Journal of Science, Engineering and Technology Research (IJSETR), Volume 3, Issue 5, May 2014
ISSN: 2278 – 7798 All Rights Reserved © 2014 IJSETR 1532
Fig -4.4: Retrieval Results of Elephant
V. CONCLUSION
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). Similarity between the images is ascertained
by means of a distance function. The proposed method
outperforms the other retrieval methods in terms of Average
Precision. Moreover, the computational steps are effectively
reduced with the use of Wavelet transformation. As a result, there is a substational increase in the retrieval speed. The
whole indexing time for the 1000 image database takes 5-6
minutes.
One limitation in our work is that the color feature is not enough to represent the image and use it for similarity
matching. There are some retrieved images which are not
similar to the query .The proposed system matches the
images if the dominant color is similar. We can overcome
this limitation by using more than one feature to represent
the image.
REFERENCES
[1] P. S. Suhasini, K. R Krishna and I. V. M. Krishna, “CBIR Using Color
Histogram Processing”, Journal of Theoretical and Applied Information
Technology, Vol. 6, No.1, pp-116-122, 2009.
[2] R. Chakarvarti and X. Meng, “A Study of Color Histogram Based
Image Retrieval”, Sixth International Conference on Information
Technology: New Generations, IEEE, 2009.
[3] X. Wan and C.C. Kuo, “Color Distrbution Analysis and Quantization
for Image Retrieval”, In SPIE Storage and Retrieval for Image and Video
Databases IV, Vol. SPIE 2670, pp. 9–16, 1996.
[4] M. Swain, D. Ballard, “Color indexing”, International Journal of
Computer Vision, 7, pp-11–32,1991.
[5] J. M. Fuertes, M. Lucena, N. P. D. L Blanca and J. C. Martinez, “A
Scheme of Color Image Retrieval from Databases”, Pattern Recognition
Vol. 22, No. 3, pp- 323-337, 2001.
[6] J.R. Smith and S. Chang, “Transform Features for Texture
Classification and Discrimination in Large Image Databases. Proceeding”,
IEEE International Conference on a Image Processing, Vol. 3, pp-407-
411, 1994.
[7] M. A. Stricker and A. Dimai, “Color indexing with weak spatial
constraints,” In Proc. of the SPIE conference on the Storage and Retrieval
for Image and Video Databases IV, pages 29–40, San Diego, CA, USA,
February 1996.
[8] R. Stehling, M. Nascimento, and A. Falcao, “On shapes of colors for
content-based image retrieval,” In The ACM Multimedia Conference, Los
Angeles, USA, Nov 2000.
[9] Yousef Mahdy, Khaled Shaaban, and Ali Abd El-Rahim, “Image
Retrieval Based on Content,” GVIP Journal, Volume 6, Issue 1, July 2006.
[10] C.H. Lin, R.T. Chen and Y.K. Chan, “A smart content-based image
retrieval system based on color and texture feature”, Image and Vision
Computing vol.27, pp.658–665, 2009.
[11] J. L. Shih and L. H. Chen, “Color image retrieval based on primitives
of color moments,” in Proceedings of the Vision, Image, and Signal
Processing, IEE, Vol. 149, No. 6, 2002, pp. 370-376.
[12] M. Mustaffa, F. Ahmad, R. Rahmat, and R. Mahmod, “Content-Based
Image Retrieval Based on Color-Spatial Features,” Malaysian Journal of
Computer Science, Vol. 21(1), 2008.
[13] D. Zhang, “Improving Image Retrieval Performance by Using Both
Color and Texture Features,” In Proc. of IEEE 3rd International
Conference on Image and Graphics (ICIG04), Hong Kong, China, Dec.18-
20, 2004, pp.172-175.
[14] H. Kekre, and D. 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.
[15] N. Sharma, P. Rawa, and J. Singh, “Efficient CBIR Using Color
Histogram Processing,” Signal & Image Processing : An International
Journal(SIPIJ) Vol.2, No.1, March 2011.
UdayaTheja.V, M.Tech(CNE), Dr.AIT, Bangalore, India1
Sangamesh, M.Tech( CNE), Dr.AIT, Bangalore, India 2
Dr.Rajshekhar Ghogge, Assistant Professor, Dept of ISE,
Dr.AIT, Bangalore, India3