Bonfring International Journal of Software Engineering and Soft Computing, Vol. 6, Special Issue, October 2016 213
ISSN 2277-5099 | © 2016 Bonfring
Abstract---- In today’s world, technology is enhancing day
by day, the most enhanced research area in digital image
processing is image retrieval system .The techniques used for
retrieving image on the basis of sketches that can be describe
as an image. So the system is referred to as Sketch Based
Image Retrieval System (SBIR). Here we made the comparison
of efficiency of different methods like Manhattan, Euclidean,
Correlation, Spearman and city block distances approaches
on different dataset and we find which is the best method
based on the efficiency.
Keywords--- Classification, Sketch, 3D Model, 3D
Retrieval, SBIR.
I. INTRODUCTION
ith the increasing number of 3D models created every
day and stored in databases, the development of
effective and scalable 3D search algorithms has become an
important research area. Generally speaking, their objective is
to retrieve 3D models similar to a 2D/3D sketch/image or a
complete 3D model query from a large collection of 3D
shapes
Query-by-Sketch (sketch-based) 3D retrieval is to retrieve
a list of 3D models that closely match a provided input
sketch. It is more challenging because of the semantic and
representational gap between the 2D query sketches and the
3D sketches, and because user sketches may vary widely in
sketching style and level of detail, as well[4][2][3]. It has
many applications, including sketch-based modeling and
recognition, and sketch based 3D animation.
Sketch-based 3D model retrieval is focusing on retrieving
relevant 3D models using sketches as input[1-5].The sketches
are Hand-Drawn sketches as shown in the figure[1]. For
objective evaluation, we have collected a large number of
query sketches .
Poornima Raikar, HOD, Dept. of CSE, KLS’s VDRIT, Haliyal,
Karnataka, India.
Dr.S.M. Joshi, Dept. Of Computer science, HOD, SDMCET, Dharwad
Karnataka, India. E-mail:joshshree@yahoo.co.in
DOI:10.9756/BIJSESC.8280
Figure 1: Images Stored in the Database with 5 Classes
(a)Planes (b)Beds (c)Bicycles d) Bees e)Barns
An image is retrieved from the database in several ways in
user queries. SBIR is one of the efficient and important
methods which are not necessary to have a high skill to draw
the query sketch. First, we review the feature extraction,
features based matching, and indexing which represents the
base of recall images.. In this paper we discuss about how to
retrieve the images using shape , texture descriptor, then
method for retrieval, then experimentation results of methods
and lastly comparison of efficiency of distance methods.
Figure 2: Diagram for Image Retrieval
The steps involved in shape based image retrieval Query as
shown in Fig 2. Image Database contains set of sketch images
into the database. First we extract the features of the input
Efficiency of Methods in Retrieval Using Query by
Sketch
Poornima Raikar and Dr.S.M. Joshi
W