[attachment=12230]
Presented By:
Jaya Krishna.P , Prem Kumar.M
Nimra College Of Engineering &Technology

---

Abstract



Interest in the potential of digital images has increased enormously over the

last few years, fuelled at least in part by the rapid growth of imaging on the World-Wide

Web. Users in many professional fields are exploiting the opportunities offered by the

ability to access and manipulate remotely-stored images in all kinds of new and exciting

ways . However, they are also discovering that the process of locating a desired image in

a large and varied collection can be a source of considerable frustration.The problems of

image retrieval are becoming widely recognized, and the search for solutions an

increasingly active area for research and development.
Fuzzy Logic is a departure from classical two-valued sets and logic, that uses "soft" linguistic (e.g. large, hot, tall) system variables and a continuous range of truth values in the interval [0,1], rather than strict binary (True or False) decisions and assignments. Formally, fuzzy logic is a structured, model-free estimator that approximates a function through linguistic input/output associations. Fuzzy rule-based systems apply these methods to solve many types of "real-world" problems, especially where a system is difficult to model, is controlled by a human operator or expert, or where ambiguity or vagueness is common. A typical fuzzy system consists of a rule base, membership functions, and an inference procedure.

So, we have chosen these topics,to model a real world system that can

retrieve images from a large collection of data using fuzzy logic techniques

In this paper, an intelligent content-based image retrieval system is proposed.
 The system is based on neuro fuzzy technique.
 The CBIR system will accept multiple queries as input that can be provided on-line and the outputs of the system are the images with their confidential values.
 The system uses fuzzy logic to interpret multiple natural expressions.
 The system can be implemented on World Wide Web.
 This paper exemplifies CBIR using an ‘intelligent on line system’







Introduction


Due to the advent of cheaper storage devices, fast computer and

communication technologies, large collections of images is growing rapidly. Retrieving

images from large and varied collections using image content as key is a challenging and

important research . Several systems have been developed so far. These systems uses

feature extraction from the image such as color, texture and shape. These features are

usually extracted from images and stored into the databases .

Color is one of the most significant features for visual recognition and discrimination. The inherent features of the content-based image retrieval systems are "imprecision", "partial information", and "user preferences”. To retrieve the images that have combination of blue sky and green lawn in them from a database, the user has to select the colors from pallet and ask the system to retrieve the images with the matching color. The other features of the image such as texture or shape can be used similarly. This is the partial and imprecise query. From the user's point of view the given color pallet represents the part of the user's information need. From the database point of view the query is imprecise. The user can pose a query to retrieve the images by selecting green color with preference 0.8 and blue color preference 0.4. The color extraction algorithm identifies arbitrarily shaped regions with images that contain colors participating in specific color sets derived from color pallet. The index for the database is built directly from color extraction of images and it searches for color set.
It is experimentally found that the user can recognize a small set of colors.
So taking into consideration the human perceptual range, the prototype built up

distinguishes between nine colors only.

An intelligent on line system for content based

image retrieval is efficient and user friendly. The system is built using CGI script in C language. The paper is organized as follows. Neuro-fuzzy content based image retrieval technique, an intelligent on-line CBIR system is discussed, results and conclusion is given below. The effectivenes of a CBIR system depends on the query. The query is classified into three levels
 Level-1
 Level-2
 Level-3
Level 1:- Comprises retrieval by primitive features such as colour, texture, shape or the spatial location of image elements. Examples of such queries might include “find pictures with long thin dark objects in the top left-hand corner”, “find images containing yellow stars arranged in a ring” – or most commonly “find me more pictures that look like this”. This level of retrieval uses features which are both objective, and directly derivable from the images themselves, without the need to refer to any external knowledge base.
Level 2 : Comprises retrieval by derived (sometimes known as logical) features,
involving some degree of logical inference about the identity of the objects depicted in
the image. It can usefully be divided further into:
1. retrieval of objects of a given type (e.g. “find pictures of a double-decker bus”);
2. Retrieval of individual objects or persons (“find a picture of the Eiffel tower”).
To answer queries at this level, reference to some outside store of knowledge is normally required
Level 3:-Comprises retrieval by abstract attributes, involving a significant amount of high-level reasoning about the meaning and purpose of the objects or scenes depicted. Again, this level of retrieval can usefully be subdivided into:
1. retrieval of named events or types of activity (e.g. “find pictures of Scottish folk dancing”);
2. retrieval of pictures with emotional or religious significance (“find a picture depicting suffering”).
Success in answering queries at this level can require some sophistication on the part of the search. Complex reasoning, and often subjective judgement, can be required to make the page link between image content and the abstract concepts.
The most significant gap at present lies between levels 1 and 2. Many authors refer to levels 2 and 3 together as semantic image retrieval, and hence the gap between levels 1 and 2 as the semantic gap .


Neuro-Fuzzy Content Based Image Retrieval Technique


This section explains the neuro-fuzzy content-based image retrieval

technique in two stages.






 Stage 1: Referring to Figure 1, the query to retrieve the images from database is prepared in terms of natural language such as mostly content, many content and few content of some specific color. Fuzzy logic is used to define the query. We define nine colors that fall with in the range of human perception. The feature representation set of colors is: rep (color) = {red, green, blue, white, black, yellow, orange, pink, purple}. These nine colors are used as input to the neural network and the content type as output. Mostly, many and few indicate the output.






 Stage 2: Figure 2 gives the basic idea of our neuro-fuzzy content-based

Image retrieval system. Different images are downloaded from World Wide Web

(WWW) and stored as image database. Features of images such as color, texture and

Shape are extracted from the image and stored into the database. Let Fs denote the set of

all features used to represent an image;

Let Fs = {color, texture, shape}.

In the initial prototype implementation, Fs = {color}.

‘ Fu ‘ denote the set of feature values that and user can recognize. Feature representation set is motivated by the research results that can identify nine colors that fall within the range of human perception. The color feature is extracted from the images using some programs. These colors are stored in a separate database as the feature database.

Users can provide the queries in terms of natural language

such as mostly, many and few. So assume the particular color content for each image

to be “mostly”, “many” and “few”. In this model, the interpretation domain is a fuzzy set.

The ranges of the values used in the Fuzzy set are:
• [0.9, 1] for “mostly”,
• [0.4, 0.5] for “many” and
• [0.15, 0.25] for few.


Also, the numeric weights such as 0.9 and 0.92 are so close to each

other that they both indicate the particular feature is mostly present in the desired images.

• Fu is defined as the set of nine colors.

• Np represents the number of pixels in each image.

For each color value (red, green, blue, etc.) we record the number of the pixels that

belong to the value, denoted by Nf where f Fu. The output obtained after training the

neural network is used to calculate the confidential factor for each image for the specific

query.
Intelligent On-line CBIR System


The intelligent on-line system with multiple queries is

based on techniques presented in the previous sections. The overview of the system is

shown in figure 3. The input is provided through web graphical user interface that has

two parameters: color (red, green, etc.) and content (mostly, many and few). The system

takes into consideration all input provided and then it takes the relevant images from

Database/web one by one and outputs the confidential values for all selected images. The

image with highest values means that the image is most suitable to the particular query.

All the images are displayed by confidential values in descending order.

Images 1 (0.99)

Query 1 Mostly Red Image 2 (0.98)

Query 2 Many Green Images 3 (0.89)






On input query of “Mostly Red & Many Green” the search

will be performed in following way:
o First all images with “Mostly Red” query will be searched
o Next search, i.e. “Many Green” will follow on the images found by previous query


Results


A number of experiments were conducted to test the

Effectiveness of the system. A collection of the images from World Wide Web

(WWW) as a database. These images are loaded into our system. If the user wants to

retrieve the images from database such as" Mostly Green and Few Yellow", the two

colors are selected from color pallet one after the other. Similarly content type such as

Mostly, Few are selected from tabular form. Natural language is used for the query rather

than giving in numerical form such as 50% or 70% etc.


CBIR System Based on Neuro-Fuzzy Techniques






Figure 4 shows the result of the query along with the colors and the content

types specified for the query. The images obtained along with their image number and

confidential factor. These images are mainly of green leaves with yellow flowers. The

content of green color isto maximum extent with few yellow colors. In figure 4, Image

#72 contains good combination of mostly green and few yellow from all the images from

our database, so this image has maximum confidential factor, 0.992769. All other images

are obtained in descending order with their confidential factor.





The results obtained after submitting the query "Mostly Red

and Few Green" are shown in figure 5. These images contain two colors, red to maximum

extent and few green colors. The images are searched which contains maximum red color

and after getting the images, which contains red colors again the images are selected

which contains few green color. The image # 10 has the greatest confidential factor,

0.994928. All the images are obtained with their confidential factor in descending order.

In the same way, we have performed experiments for different combinations of colors

and content types. If there are no images in the image database for the specific query

given by the user, then the user gets the message; "There are no images for the specific

query."

Applications
 Crime prevention
 The military
 Intellectual property
 Architectural and engineering design
 Fashion and interior design
 Journalism and advertising
 Medical diagnosis
 Geographical information and remote sensing systems
 Cultural heritage
 Education and training
 Home entertainment
 Websearching




Current Trends


Research topics receiving substantial attention at present include:
• improved methods for Web searching, allowing users to identify images of interest in remote sites by a variety of image and textual cues
• improved video retrieval techniques, including automatic segmentation, query-by-motion facilities, and integration of sound and video searching
• better user interaction, including improved techniques for image browsing and exploiting user feedback
• automatic or semi-automatic methods of capturing image semantics for retrieval.

Conclusion


Multiple queries based system on color and content can be

very useful content-based query tool for users of image databases on the World Wide

Web. We can extend our system to search for images using multiple queries. This

system allows the user to specify colors and content types. We can conduct a number

of experiments and the results show that the system can be used on-line for the images

on the World Wide Web. Only one neural network is used to train the query, for color

and content type. In future, work can also be incorporated with other feature of image

such as texture and shape.


Let’s hope for a day when we get our required images

within a split of a second exploiting fuzzy logic to the complete extent.




References

 [1] Kulkarni, S. Verma, B., Sharma. P, and Selvaraj, H. Content Based Image Retrieval Using a Neuro-Fuzzy Technique, IJCNN’99, Washington, 1999.

 [2] Ma, W. Y., and Manjunath, B. S. Netra: A Toolbox for Navigating Large Image Databases, Proceedings of the IEEE International conference on Image Processing, 1997.

 [3] Smith, J. R., and Chang, S. F. Single Color Extraction and Image Query, International Conference on Image Processing, Washington D.C., 1995.

 [4] References from IEEE papers