This document describes Image Seeker, an image retrieval system that allows users to search for similar images by inputting a query image. Image Seeker uses shape context and SIFT descriptors to represent and match images. It compresses image representations using deep autoencoding to greatly improve storage and search efficiency. To rank search results, Image Seeker semantically interprets the query image and performs median filtering on the distance of retrieved images from the query. Image Seeker was developed to enable searching large image collections in applications like trademarks, art galleries, retail, fashion, interior design, and law enforcement.

1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 10 | Oct 2018 www.irjet.net p-ISSN: 2395-0072
© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1754
IMAGE SEEKER:FINDING SIMILAR IMAGES
Veer Priyanka Raju1, Jadhav Nikita Sunil2, Kore Shubhangi Vitthalrao3, Pawar Pooja Somnath4,
Prof. Gawade jawahar S5
1,2,3,4,5First Author Dept. of Information Technology SVPM'S College of Engineering. Malegaon (bk)-413115, Tal.
Baramati, Dist. Pune India.
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Searching could be a necessary tool for
managing and navigating the huge amounts of knowledge
obtainable in today's modern era. whereas new looking
ways became more and more widespread and reliable in
recent years, like image-based looking, these ways could
also be a lot of restricted than text-based means that in this
they are doing not enable generic user input. Image-based
looking could be a technique that enables users to draw
generic search queries and come similar drawn pictures,
giving a lot of user management over their search content.
During this, we tend to gift Image-seeker, a system for
categorization and looking across an out-sized range of
images quickly supported their similarity. Image-seeker
introduces a method for categorization images in
extraordinarily compressed representations that permits
for quick, correct retrieval increased with a construction
ranking system. Image-seeker was tested on an out-sized
set of Images against existing Image similarity metrics, and
it shows important enhancements in terms of storage needs,
speed, and accuracy.
Key Words: Image-Seeker, PCA, LDA, Fisherface.
1. INTRODUCTION
Humans can render any object on any kind of surface
using image, and with touch devices becoming an
increasingly integral form of communication, it is
important that imagebased systems be applied to more
domains. Searching is one domain where image is a
relatively new form of input. The amount of data to search
through increases every day, and with users wishing to
search in more and more modalities, image-based
searching could offer a fast and flexible solution using
simple drawings like those. However, efficient searching
requires small indices that can deliver speed and
accuracy. While much exploration has been done toward
efficient image-based searching in recent years, image
retrieval is a relatively new field that has widespread
applications. One example could be searching for clip art
based on a image. In this, we describe image Seeker, a
image retrieval engine that combines state-of-the-art
techniques to achieve fast, accurate, and ranked search
results.
Image Seeker introduces a method for building a image
search index that is extremely compact but still
sufficiently representative of the full content to achieve
high accuracy. Furthermore, it uses these “image
signatures” to perform rapid searches over a large
number of images that are boosted by object labelling to
provide some semanticoriented retrieval. Because Image
Seeker is a modular system that combines multiple
techniques, it has two significant contributions to the field
of image retrieval: 1) enormously-compressed,
representative “image signature” embedding’s for index
searching and 2) a components-based framework for
retrieval that allows extensible stages like ranking, which
we use to implement an interpretation of semantics in this
paper to boost performance. Image-Seeker is a single
system that combines three distinct subsystems. The first
such subsystem is the image-indexing stage. In this stage,
we describe a method for optimizing storage of images in
a database using a highly compressed, searchable
representation. The image descriptors, which consider
both shape and structure, along with the compression
performed using a deep auto encoder architecture,
improve the retrieval in terms of time and space
complexity when compared to other known methods. The
second subsystem is query retrieval. Similar images are
returned based on their shape and structure according to
their distance from the query shape in a kD-tree, as bound
by an empirical threshold. The final component is a
ranker, which sorts retrieved images based on two layers
of filtering. First, semantic filtering is performed on the
search results using a support vector machine (SVM)
classifier that returns the most likely label for each image.
After the filtering based on most likely meaning is
complete, a median filter is used to eliminate any outliers
before returning the result set.
1.1 Basic Theory
Inexpensive image-capture and storage technologies have
allowed massive collections of digital images to be created.
However, as a database grows, the difficulty of finding
relevant images increases. Two general approaches to this
problem have been developed, both of which use metadata
for image retrieval:
• Using information manually entered or included in
the table design, such as titles, descriptive
keywords from a limited vocabulary, and
predetermined classification schemes
• Using automated image feature extraction and
object recognition to classify image content that is,
using capabilities unique to content-based
retrieval
With Visual Information Retrieval, you can combine both
approaches in designing a table to accommodate images:
use traditional text columns to describe the semantic
significance of the image (for example, that the pictured
automobile won a particular award, or that its engine has
six or eight cylinders), and use the Visual Information
Retrieval type for the image, to permit content-based
queries based on intrinsic attributes of the image (for

2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 10 | Oct 2018 www.irjet.net p-ISSN: 2395-0072
© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1755
example, how closely its color and shape match a picture of
a specific automobile).
As an alternative to defining image-related attributes in
columns separate from the image, a database designer
could create a specialized composite data type that
combines Visual Information Retrieval and the appropriate
text, numeric, and date attributes.
The primary benefit of using content-based retrieval is
reduced time and effort required to obtain image-based
information. With frequent adding and updating of images
in massive databases, it is often not practical to require
manual entry of all attributes that might be needed for
queries, and content-based retrieval provides increased
flexibility and practical value. It is also useful in providing
the ability to query on attributes such as texture or
structure that are difficult to represent using keywords.
Examples of database applications where content-based
retrieval is useful -- where the query is semantically of the
form, "find objects that look like this one" -- include:
• Trademarks and copyrights
• Art galleries and museums
• Retailing
• Fashion and fabric design
• Interior design or decorating
• Law enforcement and criminal investigation
For example, a Web-based interface to a retail clothing
catalog might allow users to search by traditional
categories (such as style or price range) and by image
properties (such as color or texture). Thus, a user might
ask for formal shirts in a particular price range that are off-
white with pin stripes. Similarly, fashion designers could
use a database with images of fabric swatches, designs,
concept Imagees, and finished garments to facilitate their
creative processes.
2. LITERATURE SURVEY
a) Specifying gesture by example: Gesture based
interfaces offer an alternative to traditional
applications (and many others) the module that
distinguishes keyboard, menu and direct
manipulation interfaces. The between the gestures
expected by the system, known as the ability to
specify objects, an operation, and additional
pagesture recognizer, is hand coded. This code is
usually parameters with a single intuitive gesture
appeals to both complicated, making the systems
(and the set of gestures novice and experienced
users. Unfortunately, gesture-based accepted)
difficult to create, maintain, and modify. Interfaces
have not been extensively researched, partly
becreating hand-coded recognizer is difficult. This
is
one cause they are difficult to create. This
describes reason why gestural input has not
received greater attend-GRANDMA, a toolkit for
rapidly adding gestures to direct. This describes
how gesture recognizes may manipulation inter-
faces. The trainable single-stroke be created
automatically from example gestures, removing
gesture recognizer used by GRANDMA is also
described.[1]
b) Mechanix:A natural Image interface tool for
teaching truss analysis and free-body diagrams
Using artificial intelligence, Mechanix can
determine not only the component shapes and
features of the diagram but also the relationships
between those shapes and features. It is very
difficult to search images similar to a complex
scene. [2]
c) Image Classification and Classification-driven
Analysis using Fisher Vectors Provides a
classification-driven analysis which is able to
recover semantic aspects of the individual
Imagees, such as the quality of the drawing and
the importance of each part of the Image for the
recognition.[3]
d) MindFinder: Image Search by Interactive Imageing
and Tagging It is a bilateral interactive image
search engine. It enables multiple actions for users
to flexibly design their queries in a bilateral
interactive manner by leveraging the image
collection in real time. It is really not easy to
search images similar to a complex scene.[4]
3. SYSTEM DESCRIPTION
This project mainly used for Image seeker System of
new or old student’s and college department. This
project we use many operations for keeping record.
We use student registration and college department
record stored and performs for the HOD of
Department form, search form and their College and
department, student’s record. In this project we need
to fill up the basic information about the College
Management System into the registration form, total
qualifications, percentages, result, and all of college
record, and student and staff Employee current &
permanent address etc.
Two different users will be using this product:
• Admin who will be acting as the
administrator.
• Staff of the University who will be accessing
the software.
The features that are available to the Admin are:
• An Admin can Search the student
• Can view/Add the different Branches of
University available
• Can view the List of Student available in each
Branch
• Add Student and their information of the
student to the database

3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 10 | Oct 2018 www.irjet.net p-ISSN: 2395-0072
© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1756
• Edit the information of the existing Student.
Can check the report of the issue of Student.
• Can access all the accounts of the Student.
The features available to the Staff are:
• Can view the different Branches available in
the University
• Can view the List of student available in each
branch
• Can own an account in the University
• Can view the student of branch allocated to
him
• Can put a request for a new student search
• Can search for a particular Student
Figure 1 Architecture of Image-Seeker
4. APPLICATIONS
• Trademarks and copyrights
• Art galleries and museums
• Retailing
• Fashion and fabric design
• Interior design or decorating
• Law enforcement and criminal investigation
5. CONCLUSION
An image retrieval system for finding, information similar
to a given query image. We use both shape context and
SIFT key-point descriptors in the matching framework.
These Image representations are then heavily compressed
using deep auto encoding and stored in a PCA for
enormous improvements in storage and speed efficiency.
Finally, we rank the result set retrieved for an input image
by the semantic interpretation of the query paired with
median filtering on the distance of the matches to the
query image.
REFERENCES
[1] Seth Polsley, Jaideep Ray, and Tracy Hammond,A.
ImageSeeker: Finding Similar Imageesr, IEEE, 2017.
[2] O. Wobbrock, A. D. Wilson, and Y. Li,B. Gestures
without libraries, toolkits or training : A recognizer for
user interface prototypes, in Proc. , 20th Annu. ACM
Symp. User Interface Softw. Technol., New York, NY,
USA: ACM,
2007
[3] T. Y. Ouyang and R. Davis, ChemInk A natural real-time
recognition system for chemical drawings, in Proc 16th
Int. Conf. Intel. User Interfaces, NewYork,NY, USA:ACM,
2011
[4] S. Valentine et al., Mechanix A Image-based tutoring
system for statics courses, in Proc 24th Innovative
Appl. Artif. Intel. Conf., 2012
[5] D. Rubine, Specifying gestures by example, Comput.
Graph., vol. 25, pp. 329337, 1991. 25