IRJET- A Review on Automated Systems for Deformation Detection in Glasses

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 12 | Dec 2018 www.irjet.net p-ISSN: 2395-0072
© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1592
A Review on Automated Systems for Deformation Detection in Glasses
Sruthi R1, Praveen A2, Rajesh Kumar K N3, Indhu R4
1,2,3,4Dept. of Computer Science and Engineering, KPR institute of Engineering and Technology, Tamil Nadu, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract – Glass defect inspection is a tedious task to
accomplish because of its brittleandtransparentnature.
The coatings and types of glasses increases the need to
provide different mechanisms accordingtotheirtexture.
In many glass industries the quality inspectionofglasses
is done manually with a helpofhumanexperts.However,
automatic inspection should be employed in order to
improve the accuracy of defect detection as manual
techniques may increase the risk of false alarms
according to psychology of the examiner. In this review,
proceeding from different researchesinthefield,various
machine vision systems and laser-based systems is
elaborated which can be used for preferred inspectionin
glass industries.
Key Words: glass products, defect detection, machine-
vision system, laser-based system, image processing
1. INTRODUCTION
Glass surfaces aregenerallycategorizedintoninetypes
with respect to various ingredients and additions
present along with them. They are soda glass, colored
glass, plate glass, safety glass, laminated glass, optical
glass, Pyrex glass, photochromatic glass and lead
crystal glass. The deformations in them may be
scratches, dusts, foreign material, lowcontractdefects,
spots, holes, bubbles and dirt. The automatic defect
detectionmechanisms have greatly reduced theriskof
manufacturing damaged products. The application of
the system has made a great impact in customer
satisfaction and procuration of mentioned industrial
glass products in the market. The various methods
provide an effective method to determine the
detrimentsintheglasses.Furthermore,theseprevalent
technologies have grown rapidly overtime and
provides an insighttonumberofproblemsariseduring
the inspection ofglassmaterialsandremedytoremove
them. As known, Quality inspection is the important
pace of the glass manufacturing and they act as a basic
bridge between all B2BaswellasB2Ccommunications
and it take a great deal to classify the various types of
defects.Theproposedworkhaselaboratedthesections
of machine vision systems and laser based systems by
various researchers. However, surface may differ and
so the technique for inspection will also differ.
Therefore, the objective survey aims to provide
information about various machine-vision systems
using a variety of technologies that are prevalentinthe
market.
2. AUTOMATED VISUAL INSPECTION
The glass inspection systems are, equipped with all
commonly used interfaces for offline and inline use at
every step of the production processchain,notonlyfor
quality control, but also for a means of optimizing the
processes themselves. Machine vision (MV) is the
technology and methods used to provide imaging-
based automatic inspection and analysis for such
applications as automatic inspection, process control,
and robot guidance,usuallyinindustry.Incombination
machine vision systems and laser based system for
glass inspection provides a robust, precise and
accurate detection of deformations at the surface of
number of glass materials
2. SURVEY
2.1 Laser Based Systems
In 2009,[1] Mr.Ono et al proposed a system for
detecting structural defects in synthetic silica glass. It
uses an ArF/KrF excimer laser to obtain sample along
with a spectrometer working at an X-ray frequency.
Electron multiplier charged couple device, and a
monochromator are used to capture the
photoluminescence. The electron spin resonance and
photoluminescence are compared with the absorption
coefficient. The fabrication measures are consolidated
and plotted as a curve for easier reference. However,
the system needs to be understood in detail in order to
detect the surface deformations of the glass.
In 2010,[2] Mr ono et al also proposed a system for
Laser inducedStructuralDefectsinHighlyTransparent
Synthetic Silica glasses which address the problem of
optical durability with the aid of in-situ
photoluminescence (in-situ PL) and differential
absorption(in-situDA)measurementsystemstoreveal
the defects repaired by ions involved in the process
such as hydride or hydroxide. The creation and
anhelation mechanisms proposed inthesystemactsas
an important breakthrough in order to produce silica
glass with improved optical durability. But, its

necessary to understand the scenario and set up the
parameters according to our inspection criteria
2.2 Machine Vision based systems
In 2009,[3] Francesco Adamo et al uncovers a
prototype for automatic glass inspection system using
electronics and two CCD area scan camera. The major
steps provided by the system viz., calibration of
camera, reference image acquisition and analysis of
defects are satisfying the fundamental aspects of the
system. For instance, they have considered an
inspection technique for satin glass along with image
processing and customised user interface that could
uplift the knowledge about defect to the person who is
examining the product. Thus, the system provides a
reliable and cost-effective solution for the glass
industry. At the same time the position of the glass in
the inspection module must be taken into account
before examination.
In 2013,[4] Jaina George et al proposed a system with
three configurations namely feeding unit, vision
processing and sorting unit. The method detects small
defects in non-uniform intensity and low contrast
images and it is rather robust with respect to changes
of the glass type and to other operating conditions.
Image filtering is done by applying high pass, low pass,
median, Weiner and gaussian filters. The filtering
provides an enhancement to the image and improves
its accuracy. The primary objective of the system is
achieved by using fuzzy logic in order to determinethe
results from unsupervised data. The PSNR value
provides clarity obtained by various filters and
compared. Thus, the correlation between original and
noisy image from the data is addressed in the system
which automatically improves its efficiency.
In 2016,[5] Ming Chang et al constructed an optical
inspection platform for surface defect detection in
touch panel glasses. The main components of the
machine vision system are a set of hardware and
software components intended to produce high
performance computing. For instance, in order to
achieve high speed processing, the system utilizes a
GPU computing setup along with CPU for improved
efficiency. The illumination source which contains an
LED acts as a corner stone in the identification of
defects. The input raw image is subjected to histogram
equalization, binarization and labelling of images
which are a part of image processing techniques. The
speed of examination is extremely high and the
labelling is done more accurately. To recapitulate the
system provides a far better speed, accuracy and
labelling rather than any other systems taken into
examination.
In 2017,[6] Fu Li et al proposed a detection method
based on connectivity domain characteristics of the
glass bottle was presented for glass bottle defects with
a help of machine vision system accompanied by ring
led setup, camera and equipment for performing
examination in bottles. The pre-processing eradicates
the noise in image during acquisition and gaussian
filter is applied to provide more insights in the image
data. The system infers three deformations namely big
crack, small crack and bubble treatment in order to
enrich the quality of the glass bottles. They have
achievedabsolutelyhighaccuracyininspectionmaking
the visual inspection system a truly economical
machine.
In 2018,[7] Jia li et al formulated a machine vision
detectionmethodforautomatic,real-timeandaccurate
realization of glass scratch defect analysis and
qualification, which has marked its accuracy in robot
engineering and its applications. The CCD camera is
used to capture the illuminated image of the glass with
the help of LED lights embedded in the system. The
image processing undergoes thresholding, vote
detection and edge detection in the machine. The
processed image is subjected to finding the
measurement of width and height of the scratch using
the below formula
And the qualification criteria can be fixed by the
examiner. Thus, the system provides both practical as
well as theoretical values that can be used for
inspection by the system.
3. CONCLUSION
On comparing the laser based and machine vision systems
for inspection the laser system involves a large set of
parameters to be understood and applied accordingly,butin
case of machine vision systems they are more robust, more
accurate and can be applicable to variety of glass products.
The paper aims to provide a generalized overview of about
seven machines that are prototyped, as well as marketed in
order to provide people with more flexibity of choices. All
the machines used for defect detection varies certainly in its
visualization method, construction and plethora of
parameterising. In Future more applications involvingother
materials such as metals can be surveyed in order to
produce extensive study details about the defect detection

REFERENCES
[1] Jia Li, Fei Zhao, Tengfei Zhang,HuihuiMiao “Researchon
Robot Vision Detection Method for Scratch Defects of
Flat Glass Based on Area Array CCD “ 2018 3rd Asia-
Pacific Conference on Intelligent Robot Systems
[2] Fu Li , Zhou Hang , Gong Yu, Guan Wei1,Chen Xinyu1
”The Method for Glass Bottle DefectsDetectingBasedon
machine vision” 2017 29th Chinese Control And
Decision Conference (CCDC)
[3] Ming Chang, Bo-Cheng Chen, Jacque Lynn Gabayno &
Ming-Fu Chen “Development of an optical inspection
platform for surface defect detection in touch panel
glass” 2016 in International Journal ofoptmechatronics,
[4] Jaina George, S. Janardhana , Dr.J.Jaya Akshaya,
K.J.Sabareesaan “Automatic Defect Detection In
spectacles And Glass Bottles Based On Fuzzy C Means
Clustering” International ConferenceonCurrentTrends
in Engineering and Technology, ICCTET’13
[5] Francesco Adamo, Filippo Attivissimo, Attilio Di Nisio
and Mario Savino “An automated visual inspection
system for the glass industry ”2009 - International
Instrumentation and Measurement Technology
Conference
[6] M. Ono, A. Koike, K. Iwata1 , M. Takata1 “Observation of
ArF Laser induced Structural Defects in Highly
Transparent Synthetic Silica glass” 2010 Optical Society
of America
[7] ” M. Ono, A. Koike, T. Ogawa1 , M. Takata1 ,S.Kikugawa1
“Detection of Structural Defects of Extremely Low
Concentrations in Commercial Synthetic Silica glass”
2009 Optical Society of America
[8] Shimizu, M.; Ishii, A.; Nishimura, T. Detection of foreign
material included in LCD panels. In Proceedings of the
26th Annual Conference of the IEEE Industrial
ElectronicsSociety,Nagoya,Japan,October22–28,2000;
IEEE: New York, 2000; pp. 836–841.
[9] Zhao, J.; Kong, Q.J.; Zhao, X.; Liu, J.; Liu, Y. A method for
detection and classification of glass defects in low
resolution images. In Proceedings of the 6th
International Conference on Image andGraphics,Anhui,
China, August 12–15, 2011; IEEE: New York, 2000; pp.
642–647.
[10] Yajnavalkya Bandyopadhyay “Glass Defect Detection
and Sorting Using Computational Image Processing”
October 2015, Volume 2, Issue 10 ,JETIR (ISSN-2349-
5162)
[11] Timothy S Newman “A survey of automatic visual
inspection” Computer vision and image understanding
Vol 61,No 2, March , pp 231-262 , 1995
[12] João David Daminelli Cabral & Sidnei Alves de Araújo “
An intelligent vision systemfordetectingdefectsinglass
products for packaging and domestic use” Int J Adv
Manuf acturingTechnology DOI 10.1007/s00170-014-
6442-y

IRJET- A Review on Automated Systems for Deformation Detection in Glasses

More Related Content

What's hot (6)

Similar to IRJET- A Review on Automated Systems for Deformation Detection in Glasses (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET- A Review on Automated Systems for Deformation Detection in Glasses