Limitations of Barcode and QR code Scanner

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 09 | Sep 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1178
Limitations of Barcode and QR code Scanner
Manan Bhand, Ishaan Dwivedi
------------------------------------------------------------------------***-------------------------------------------------------------------
Abstract
Standardized identification wound up fundamental components in deals and items benefits because of significance of
tracking all things in on spot. For this reason, there are numerous strategies actualized to make the standardized tag
perusing procedure wound up simpler to clients. This venture is to build up a standardized identification acknowledgment
framework by utilizing picture preparing. The framework will have the option to peruse scanner tag through a picture and
the framework able to catch the picture by utilizing a webcam. This task will utilize Python programming project to build
up the framework and it will incorporate with webcam or computerized camera. Framework will investigate the picture
and after that show on the Graphical User Interface (GUI) the scanner tag type, information and size of the picture.
Framework is intended to perceive various sorts of standardized tag and show the information once the scanner tag
picture is caught. Framework additionally is to give comfort method for watching information from the standardized
identification with lower costing looked at by utilizing the electronic standardized tag scanners. This frameworkcan be
utilized whenever and anyplace by the client who likes to watch the information spoken to by the standardized
identification numbers without going any spots giving the scanner tag scanner administrations.
As the outcome, the undertaking has been grown easily and superbly. For the future framework improvement, it is
recommended that the framework likewise should comprise aslider with the goal that the client would ready to control
the brilliance of the picture which caught by webcam. Speedy Response (QR) codes are two-dimensional scanner tags that
canbe utilized to proficiently store modest quantity of information.
They are progressively utilized in all life fields, particularly with the wide spread of advanced cells which are utilized as QR
code scanners. While QR codes have numerous preferences that make them famous, there are a few security issues and
dangers that are related with them. Running vindictive code, taking clients' touchy data and damaging their protection and
wholesale fraud are some ordinary security chances that a client may be liable to out of sight while he/she is simply
perusing the QR code in the closer view. In this paper, a security framework for QR codes that ensures the two clients and
generators security concerns is executed. The framework is in reverse good with current standard utilized for encoding
QR codes. The framework is actualized and tried utilizing an Android-based cell phone application. It was discovered that
the framework presents somewhat overhead as far as the defer required for honesty check and substance approval.
Key words: QR code, Python, GUI, tag scanner, Android
1.Introduction
Standardized tag is a visual portrayal of data as bars and spaces on a surface. The bars and spaces are structured with
various widths and comprise of numbers, characters and images,for example, spot, colon and others. Various mixes of
these alphanumeric characters are utilized to speak to data. There are different sorts of scanner tags being used today for
example Code 128, Code 39, EAN and so forth. (Mind, 2000).
The fruitful of scanner tag innovation has been continually improving so as to suit more data in the base conceivable space.
Today scanner tags are generally utilized on books and at retail locations so as to monitor the items accessible and simple
checkoutof the items. The standardized tags are regularly perused utilizing scanners utilizing laser pillars or cameras.
By and large, standardized tags are images molded as square shapes which comprise of slim or thick parallel lines parallel
to one another. Standardized tags give intends to programmed fast information contribution to the PC. Since the most
recent decade, scanner tags are being utilized in numerous territories, for example, showcase items and electronic gadgets.
The lines on scanner tags contain the reference number of the item. This data ought to be recorded in PCs to store every
item independently for checking organization deals and buy amounts. When perusing standardized identifications on
items utilizing some laser checking gadget, a sign is produced by the framework and prepared in the PC by programming.
At thatpoint this data is utilized to figure out which item is chosen. This procedure gives fast and dependable deals chances
to organizations for selling their items. There are a few sorts of scanner tag that being utilized inside the mechanical field
these days. A standardized identification symbology characterizes the specialized subtleties of a specific sort of scanner
tag incorporates width of bars, character set, strategy for encoding and checksum determinations. Standardized

identification types can be grouped into four kind of classification and those classes are numeric-just scanner tags, alpha-
numeric standardized tags, 2D standardized identification and industry standard for standardized tag and marks.
As in picture preparing, the thresholding is the one of component to be consider as an approach to portray a picture.
Thresholding is the least complex technique for picture division. From a grayscale picture, thresholding can be utilized to
make double pictures. During the thresholding procedure, singular pixels in a picture are set apart as item pixels if the
worth is more prominent than some edge esteem (accepting an article to be more brilliantthan the foundation) and as the
foundation pixels generally. This show is known as edge above. Variations incorporate limit underneath, which is inverse
of edge above; edge inside, where a pixel is marked "object" if its worth is between two edges; and edge outside, which is
something contrary to edge inside. Normally, an article pixel is given an estimation of "1" while a foundation pixel is given
an estimation of "0." Finally, a twofold picture is made by shading every pixel white or dark, contingent upon a pixel's
mark.
2.Literature Survey
A. 2D Barcode Detection using Images for Drone-assistedInventory Management
2018 15th International Conference on Ubiquitous Robots (UR)Hawaii Convention Center, Hawai'i, USA, June 27-
30, 2018.
Hyeon Cho, Dongyi Kim, Junho Park, Kyungshik Roh, WonjunHwang
Automaton helped stock administration is alluring for organizations with enormous distribution centers and industrial
facilities. Moreover, there is an enthusiasm for a novel strategy that consequently recognizes target standardized tags
utilizing an IR-based camera, which empowers productive automaton way arranging and results in diminishing force
utilization. In this paper, we propose a proficient recognition structure which decides the restrictions of 2D standardized
tags. Numerous territorial recommendations of 2D standardized identifications are diminished to a couple of up-and-
comer areas as per the separation data between the automaton and the objective 2D scanner tag. Visual highlights of the
chose competitor areas are extricated by LBP and HOG strategies, separately. To pick up discriminant control for
characterization, SVM is utilized toward thepart of the bargain. The last discovery area is dictated by a total based score
combination technique that is weighted. To approve the presentation of this strategy, we gather 2 dimensional scanner tag
pictures under genuine distribution center conditions and acquire broadtest results.
The boundless utilization of automatons has fundamentally added to numerous applications from characteristic asset the
executives, air contamination checking, distribution center stock administration, and so forth. Among them, we have an
enthusiasm for an automaton based administration arrangement intended to computerizestock taking and stock control in
enormous scaled distribution centers. It has numerous alluring highlights contrasted and the unmanned ground vehicle
(UGV) as of now in activity. For instance, it isn't important to create new street lines for an UGV when the capacity area is
altered in the distribution center and street line is lacking since it has been exhausted in light of the numerous UGV
developments. Besides, extra hardware isn'trequired to find an objective thing found in some tall shelf. An automaton has
its very owncamera gadgets and it gets around the distribution center autonomously as per a preordained flight intend to
catch the important data on beds. Initially, we select the best districtcompetitors utilizing a particular inquiry strategy and
separate the notable highlights from the up-and-comers. Toward the end, thearrangement result is controlled by a straight
SVM for a quick estimation
In this papera framework is proposed that utilizations three stages, area up-and-comer identification, include extraction,
and SVM grouping, for standardized identification location and acknowledgment in production line stockrooms. We
connected a component calculation to four sorts of 2D standardized tags and estimated the presentation by lookingat the
accuracy and learning review by highlight. Furthermore, we additionally improved execution by means of a late
combination of the outcomes from different highlights. The proposed technique shows an exactness of 98.08% and a
review of 98.27%.
B. A self-powered 2D barcode recognition system based on sliding mode triboelectric nanogenerator for
personal identification Jie Chen, Xianjie Pu, Hengyu Guo, Qian Tang, Li Feng, Xue Wang, Chenguo Hu, 11
November 2017
Triboelectric nanogenerator (TENG) for creating self-controlled individual distinguishing proof framework displays
extraordinary points of interest, be that as it may, a few issues (swiping speed, soundness, and so on.) in the past research

work counteract this sort of standardized identification framework for future work. Thus, here, we plan a triboelectric
based scanner tag acknowledgment framework (T-BS) for down to earth applications. By utilizing a reference scanner tag
part, the yield signal under arbitrary swiping movement can be effectively perceived, which offers a phenomenal
procedure for the arrangement of the issues in the past work. Through the combination oflimit esteem, stage distinction
and crest looking, the yield voltage sign of the gadget can be changed into computerized signals "1" or "0", and the data can
be continuous shown on LabVIEW stage. When a card is swiped at a consistent speed, or non-steady speed by human
hands, the coded data can be effectively recognized and the entrance control framework works precisely. Additionally, by
expanding the quantities of standardized identification sections also, minifying the size, progressively exact 2D scanner tag
acknowledgment framework (counting scanner tag and peruser) is gotten and understood the individual recognizable
proof. This exhibits the probability of the TENG as a self-fueled 2D standardized identification acknowledgment
framework for individual recognizable proof, whichshows potential applications in data security
Keywords: 2D barcode, triboelectric nanogenerator, self-powered, recognitionsystem, personal identification
Lately, with the fast advancement of web of things (IOTs), individual data recognizable proof innovation has turned into
the extremely essential and the most significant connection of different administrations. Along these lines, endeavors have
been given to the improvement of sheltered and solid individual data security framework [1-3]. Among the present
advances, a one-dimensional or two-dimensional standardized identification [7, 8] is the most well-known and broadly
utilized methodology for datatransporter. In any case, a large portion of them are essentially coded by separating thegame
plan of spaces what's more, bars and require an outside power supply. we have exhibited a self-fueled triboelectric based
2D standardized tag acknowledgment framework for individual recognizable proof. With the presentation of a reference
standardized tag part, the coded data of data scanner tag under irregular swiping movement can be precisely perceived,
which presents an incredible answer for the significant disadvantage existing in the past work. Through the determination
of edge worth and stage contrast, the exactness of pinnacle is incredibly improved.
The yield sign of the gadget can be changed over into computerized signals "1" or "0", and the coded data is shown
progressively. At the point when a card is swiped at steady speed, or non-steady slidingrate by human hands, the coded
data can be effectively distinguished and the entrance control framework works precisely. Besides, by improving the
segments of standardized identification and minifying the size, increasingly exact 2D scanner tag is acquired and a lot
more secure individual distinguishing proof framework is accomplished. This work shows the probability of self fueled 2D
scanner tag framework dependent on the TENG to acknowledge individual ID, which shows potentialapplications in data
security.
C. Fast Real Time 1D Barcode Detection From Webcam Images Using the Bars Detection Method Proceedings of
the World Congress on Engineering 2017 Vol IWCE 2017, July 5-7, 2017, London, U.K Abderrahmane NAMANE,
Madjid AREZKI
The location of 1D standardized identification from foggy, low difference and low goals pictures is as yet a difficult issue.
1D standardized tags are comprised by various bars. In this work, another bar recognition strategy (BDM) is produced for
precise and quick 1D standardized tag identification. The least square technique is acquainted with decide precisely the
direction of the distinguished bars. Two likelihood thickness capacities (PDF) of the direction and length of bars are
utilized and connected to the distinguished bars so asto decide the predominant direction and length of the current bars.
At last, the Hough change (HT) is connected through these recognized centroids of the bars, so as to the bars thatare by one
another, and lie on a similar line bolster which compares to the scanner tag direction. Fantastic outcomes were
accomplished on still pictures from two one dimensional standardized identification datasets: WWU Muenster Barcode
Database and ArTe-Lab 1D Medium Barcode Dataset. Exploratory outcomes demonstrate that our calculation can get
better 1D standardized identification location contrasted and existing techniques. They demonstrate that the capacity of
the model yield applicable and powerful scanner tag recognition even with low goals andobscured outlines, and for bent
standardized tags.
Keywords—1D barcode detection, orientation estimation, least square method, Houghtransform
Our primary thought is to follow the bars in item picture so as to identify the 1D scanner tag. In this paper we utilize an
alternate methodology dependent on separating, binarization, external form recognition, bars discovery with BDM, LS for
bars directions,overwhelming bars direction estimation ( d) with the first PDF, predominant bars length estimation (ld)
with the second PDF and Hough change through just the prevailing bars direction ( d+90°) which means our work is
precise and exceptionally quick, and suits forongoing application. The proposed identification strategy depends basically
on the eight principle steps, performed successively. The information picture of the framework speaksto external shapes of

the caught scene. This picture is gotten in the wake of preprocessing, where the sifting, binarization and form
identification are guaranteed by Gaussian separating, nearby binarisation technique and outskirt following calculation
separately. We have displayed a 1D standardized tag recognition strategy connected to pictures from two databases and to
pictures from Webcam gadget. This particular technique and framework not exclusively can recognize standardized tags
in complex scene yet in addition it can distinguish obscured scanner tags because of unfocused Webcam. Our framework
can work progressively because of its fast, and can process a picture in 21ms. Trials were led on genuine pictures of items,
with a low goals and a poor focal point of the Webcam. We show exploratory outcomes that demonstrate a high
accomplishment on particular databases, namely, the WWU Muenster and Arte-Lab 1D standardized identification,
contrasting and the current strategies. These outcomes demonstrate the power of the proposed strategy. We show
likewise that the BDM is thickness subordinate, the more the bar is meager and the more its discovery is obvious. The
proposed technique demonstrates 100% precision continuously applications.
D. Automatic Barcode Extraction for Efficient Large-Scale Inventory Management Lichao Xu, Vineet R. Kamat, and
Carol C. Menassa
Standardized tags are widely utilized in stock administration to distinguish and follow items and segments. While the
current pervasive technique for perusing standardized tags with handheld scanner tag perusers is reasonable for retail
checkout paths and little scale stock administration, it presents noteworthy work effectiveness and specialist wellbeing
challenges for enormous scale stock administration where items are typically put away on multistory racks in huge
distribution centers or storerooms. Such offices present noteworthy chances to improve coordinations via robotizing the
stock administration process. To straightforwardness stock following while at the same time alleviating the distinguished
issues, this exploration exploits PC vision based scanner tag perusers to supplant conventional standardized identification
perusers and proposes three methods toencourage standardized identification extraction from video filter information. In
the firstplace, Harris corner identifier and Hough change are proposed to work as one to evaluatethe bearing of an area
including a solitary standardized tag and pivot it to a perfect state to serve existing unraveling calculations. At that point, a
calculation dependent on scanner tag locale network and geometry property is proposed to discover different
standardized tags in a solitary picture to maintain a strategic distance from animal power looking of substantial
standardized identifications. What's more, a histogram distinction based key casing choice technique is likewise proposed
to wipe out excess data between consecutive outlines which improves proficiency by utilizing less outlines. At last, these
systems are connected to video information gathered in a huge coordinations distributioncenter and the general execution
is assessed. The exploratory outcomes demonstrate that our calculation can accomplish both attractive acknowledgment
rates and productivity, and along these lines offers huge guarantee for wide organization in robotized huge scale stock
administration utilizing earthly and ethereal mechanical stages.
Organization of scanner tag frameworks in stockrooms and dispersion focuses has essentially improved stock
administration by permitting exact stock following and high work effectiveness. Be that as it may, directors still face
challenges in admirably using stockroom assets because of genuine insufficiencies in exceptional stock data. Such
enormous data slacks are fundamentally brought about by the current predominant however wasteful method for filtering
standardized tags with a normal (for scanner tags at low places) mounted on a forklift (for standardized identifications at
high puts) standardized identification scanners as appeared in Figure 1. While perusing standardizedidentifications with a
scanner tag scanner, standardized identifications should consistently be examined individually, requiring huge human
exertion and work everyone esaeantial entrance here walks to finish enormous scale filtering assignments in huge
distribution centers. Furthermore, specialistwellbeing is another issue generally experienced while utilizing substantial
forklifts for such reason. General techniques for perusing standardized identifications from pictures depend on perceiving
the double designs along scanlines which cross the scanner tags.
Monitoring a standardized identification's direction enormously decreases number of scanlines and increment the
possibility of effective read which can further improve perusing effectiveness and exactness. This paper proposed three PC
vision based methodsto encourage scanner tag extraction from video check information. Investigations led utilizing video
gathered at a functioning distribution center demonstrate that the calculation parts work viably to peruse out and
extricate most of the area (i.e., cell) distinguishing standardized identifications powerfully. Another essentialness of this
work isthat every one of the three procedures examined above don't utilize explicit data from different advances, which
make them simple to consolidate with different calculations or computational grouping. These qualities increment the
possibilities of their wide application, despite the fact that some specialized difficulties still stay before their viable
attainability. The primary constraint is some human exertion is as yet required for setting proper limits which can profit
by being mechanized.

E. Barcode Character Defect Detection Method Based on
Tesseract-OCR, 2017 3rd IEEE International Conference on Computer and CommunicationsGang Zhao, Luyu Lin,
Yawen Chen, Shan Liu, Jie Chu, ZhuoranLuo
With the consistent improvement of data innovation, the utilizations of standardized taghave turned out to be increasingly
more broadly, and its quality necessities are additionally expanding. Because of the destitution hardware for printing etc.,
and the defective printing innovation, there are a number of issues, for example, flying ink, missing characters, basic
errors, dark spots and inappropriate enlistment existing during thetime spent standardized tag printing. The conventional
method for physically arranging flawed standardized identification isn't just wasteful yet in addition effectively impacted
by numerous components, which prompts the low exactness of the recognition. So as totake care of these issues, this paper
proposes a technique for standardized tag deformity recognition dependent on Tesseract-OCR, initially, the strategy
utilizes the level projection strategy to fragment the scanner tag, and after that it utilizes the Tesseract OCR strategyto
perceive the characters in the scanner tag, in conclusion, it consolidates
Levenshtein Distance calculation to identify the character misses. In this paper, 1000 standardized tag pictures were
utilized to the investigation, and the and the results show that the precision of recognition results can arrive at 94.3%,
which demonstrates the possibility ofthe strategy.
As the recognizable proof characteristic of ware, scanner tag has significant research hugeness for programmed ID of
merchandise data, programmed arranging and canny deal, and so forth [1]. Be that as it may, in light of the typography that
is poor, low precision of the hardware and the blemished printing advancements, there are a ton of inquiries, for example,
flying ink, characters missing, wrong print, dark spots and ill-advised enrollment during the time spent standardized
identification printing. Conventional strategies for physically arranging inadequate scanner tags require a ton of HR, which
are wasteful, yet in addition existing emotional elements, that enormously influences the exactness of the identification. As
the standardized tag is not quite the same as the common prints, a total scanner tag can be isolated into two regions, the
character zone and the realistic identifier zone, which are remarkable in a standardized tag. The parallel preprocessing of
the standardized tag picture in the location procedure is to feature the qualities of the scanner tag characters. In the
double procedure, the limit is picked by the biggest interclass fluctuation technique (Otsu). The key advance in recognizing
the deformities ofthe characters in a standardized tag is to distinguish the characters in the scanner tag. So as to evade
each test to change the layout and to precisely find the area of the imperfection, and make the exactness of ID more
accurate, the even projection strategy is utilized toportion the standardized tag. Here, a strategy dependent on Tesseract-
OCR character deformity recognition is put forth to take care of the issue of tedious and relentless in the customary
method for standardized identification character imperfection location. By utilizing the holes between every line in the
standardized tag, the flat projection calculation is utilized to fragment the scanner tag. At that point the Tesseract OCR is
utilized to differentiate and identify the characters in the divided picture. At last, the Levenshtein Distance calculation is
utilized to identify the character absconds. The trial results demonstrate that the exactness of the technique is 94.3%. The
strategy has high acknowledgment exactness and quick speed, which demonstrates the plausibility of scanner tag
character imperfection discovery dependent on Tesseract-OCR. Be that as it may, the test set that is used here is
constrained and less quantity of test pictures are prepared by the Tesseract-OCR character for be satisfactory.
Subsequently, the relocation ofthe character library might be poor, and the exactness of the deformity discovery might be
influenced by other scanner tag characters, so extending the quantity of standardizedidentification tests is the purpose of
our next examination.
F. DEVELOPING AND IMPLEMENTING A BARCODE BASED STUDENT ATTENDANCE SYSTEM
Rahmah Al Sheikh, Raghad Al-Assami, Maryam Albahr, Muntaha Al Suhaibani, Mutasem k. Alsmadi, Muneerah
Alshabanah, Daniah Alrajhi, Ibrahim Al-Marashdeh, Sanaa Alsmadi and Hayam Abouelmagd and Mohammed
Tayfour International Research Journal of Engineering and Technology(IRJET),Volume: 06 Issue: 1 | Jan 2019
In perspective on the significance of understudies' participation at talks and their effect on their scholastic
accomplishment, colleges take the vital measures to lessen inordinate truancy. This is an exceptionally significant issue.
The organization requires cautious development, dealing with it and not being indulgent. Right now, names of the
understudies are called to mark participation and nonappearance at the colleges or by marking the understudy's
participation paper. During the time spent conceding understudies into an assessment lobby in most KSA colleges, 85% of
Keywords-Tesseract-OCR; defect detection; barcode; horizontal projection method Problem
Statement+Methodology+Expected Output

participation must be fulfilled and furthermore considered for evaluation calculation, consequently there is an enormous
requirement for observing and recording understudies' participation. The purpose of this project is to structure and
execute a scanner tag based understudy participation framework that can be effectively gotten to by the speakers, to assist
them with avoiding keeping up the library book, giving important data about the understudies and the reports can be
created utilizing constant handling. The work here was planned and executed utilizing the Unified Modeling Language
(UML), Microsoft Access 2007 and ASP.NET programming language.
Key Words: Attendance Management, Barcode Scanner, Unified Modeling Language.
It is verifiable truth that for all intents and purposes all associations whether instructive orbusiness need to appropriately
record the participation of its understudies or workers for successful arranging, the executives and working of the
association. In many colleges in the creating nations, understudy's participation is normally adopted by old record
framework strategy such as calling their names to check and utilizing paper, this methodology is being utilized for quite a
while. As indicated by Tabassam, et al., in It ends up being troublesome for the organization at the colleges to normally
refresh the participationrecord and physically ascertain the level of classes not attended and went to with the end goal of
ensuing outcomes handling and assessments. Remembering these issues, this work planned and executed a framework to
beat the issues related with participation recording. Through our perception of the present framework we found
numerous issues, including the huge lack of precision of the customary participation frameworks and the plausibility of
understudies to record their partners present when they are not present, and the time lost in this activity to call the names
of understudies to gauge participation, and the trouble of techniques to gather reasons of nonappearance, and to raise
pardons, The accompanying inquiries:
1. What advantages can be accomplished by applying electronic participation tounderstudies in colleges?
2. What is the effect of utilizing the electronic participation application on the instructive procedure of the
understudies?
3. What are the outcomes to be accomplished through the use of electronic participationin colleges?
4. By what means can employees and understudies acknowledge the technique forapplying electronic participation?
The procedure of framework examination expects to ponder a current framework to completely plan another framework.
Framework investigation is performed to accomplishfundamentally two points to be specific:
1. To comprehend the procedure or the framework unmistakably. This will aid the newframework plan.
2. System investigation will distinguish the issues in the current framework; consequently this will realize the
wastefulness reasons.
The Unified Modeling Language (UML) is representation for the framework plan, it speaks to graphical documentations
which help to portray and structure programming frameworks,primarily programming frameworks built using the item
arranged style. The
UML was used for the most part to structure the framework being proposed. The Use-Case outline and the Class chart are
tended to beneath. Automating homeroom participation following has numerous focal points over the old framework.
Information from homerooms can without much of a stretch be changed inside the databases for conceivable later
examination or uses. This work structured and executed a scanner tag based understudy participation framework that can
be easily utilized by the speakers without issue, help the instructors to abstain from keeping up the library book, giving
profitable data about the understudies and the reports can be produced utilizing continuous preparing. The proposed
framework was structured and executed utilizing the Unified Modeling Language (UML), Microsoft Access 2007 and
ASP.NET programming language. This particular framework will likewise help in creating the defaulters list individually
and send messages to those understudies that have lesser participation than is required.
G. Digital 3D Barcode Image as a Container for Data Hiding Using Steganography Rama Rani, Gaurav
Deep. 4th IEEE International Conference on Signal Processing, Computing and Control(ISPCC 2k17), sep21- 23,
2017, Solan, India.
Steganography is a method of disguising private data in a spread medium so that it ends up outlandish for the third
individual to come to realize that some secret data is present in the spread envelope. In the present time with the origin of
new rising advances, standardized identifications has turned out to be one of the most prevalent strategies to give an

instrument to securing delicate information.3D scanner tags are utilized to suit high information rates by utilizing third
measurement as a shading. 3D standardized tags fills in as the most dependable procedure to shroud information since
they don't utilize any blunder revision levels because of the fact that it is exceptionally hard to modify or change the
encoded data. Here, the idea and procedure of information covering up in standardized identifications by utilizing shading
as third measurement is presented. The procedure is arrangedinto various classes and execution is assessed by utilizing
different measurable parameters.
Keywords - Steganography; Quick Response code; Cover channel; Stego channel; PSNR; MSE, with the origin of new
correspondence ways, the requirement for giving security to thesecret information on the web is continuing expanding
step by step. There exists various data concealing applications such as watermarking, fingerprinting, copyright insurance
for advanced media and steganography. Steganography gives a way to secured composing.
Joining cryptography and steganography together results into a larger amount of data protectionand security. Advanced
pictures fills in as the most famous data conveys for inserting encoded data due to their expanded utilization on the web.
In today 'universe of systemcorrespondence, for the need of more secure methods, information can be disguised in 3D
Barcode pictures. 3D scanner tags are normally spoken to by utilizing shading as a thirdmeasurement. 3D Barcode age and
inserting stage, 3D Barcode extraction stage. In this paper, we exhibited another strategy for disguising information at all
huge bits of a hued 3D standardized tag picture. The strategy utilized, first separates red segment of the picture dependent
on some key worth. The mystery picture is changed over into encoded message by utilizing XOR encryption calculation. At
that point the least noteworthy bits ofthe separated parts are supplanted with the least critical bits of the content to be
hidden.
The methodology utilizes basic LSB (Least critical Bit) Method. The proposed plan keeps up the nature of stego picture
with a normal pinnacle sign to clamor proportion of 87.93 and Mean Square Error of 1.078 and structure closeness list
matrixof 0.999. The proposed plan expands the nature of stego picture and furthermore builds the inserting limit of 3D
hued Barcode pictures. In the foreseeable future this methodology can be bettered by utilizing a fewadvancing calculations
for upgrading the inserting limit and nature of stego picture with less contortion in the picture. A few mystery messages
can likewise be disguised in a solitary spread picture.
H. Electronic Toll Collection System Using Barcode Technology E. V. V. Hari Charan, Indrajit Pal, Akash Sinha, Raj
KamalRoye Baro and Vijay Nath
This paper accentuates the standardized identification innovation for programmed Electronic Toll Collection (ETC)
frameworks, so as to stay away from the regularly expanding stream of traffic and the long lines at the tollbooths of the
expressways. The proposed strategies utilize computerized picture handling methods to check the scanner tag and to
coordinate it with the current database, by using the decoded information. The actualized equipment arrangement has
been shown and talked about in detail. The proposed system for standardized tag identification really improves upon the
speed, the efficiency and brings down the expense of execution. The technique has been executed byutilizing Python and
OpenCV.
Keywords Barcode: QR code, RFID tag, ETC, Tollbooth
The Electronic Toll Collection (ETC) framework was created so as to defeat the issues related with the manual gathering of
the toll. The manual toll conclusion framework has been supplanted by the computerized framework, so as to stay away
from issues of long lines and blocked traffic. The computerized ETC framework enables the client to maintaina strategic
distance from the hang tight for installment of money or for gathering of token so as to cross the toll. The programmed
framework collects toll electronically through the use of RFID, standardized tag innovation, laser innovation, and so on.
With higher efficiency than the RIFD and the laser, the standardized tag innovation is talked about in detail. A parallel dark
bars and void areas arrangement of differing widths establishes a standardized identification. The components of a
scanner tag are the spaces and the bars. Various characters are spoken to by the various mixes of spaces and bars, for
example, numbers or letters. The robotized framework works quicker and all the more efficiently. It is fit for deciding the
enlistment status of a vehicle and advising experts regarding infringement in installment of toll and educating proprietor
by means of email or SMS of the area and theconcluded toll for a vehicle. The goal of this innovation is to keep tollbooths
from gettingto be clog by a vehicle line. It encourages the legislature to keep up an appropriate and a refreshed record of
the accumulation of toll. There can be a tag introduced to the front plate of the vehicle for this. This paper principally
centers around the standardized identification recognition calculation used to get the ideal working efficiency. Ideas of AI
alongside an IDE domain of Python and OpenCV have been utilized so as to get the working model. An imaging gadget like
a camera was utilized for the proper discovery of the scanner tag. The standardized tag to be examined. It stores the

financial balance subtleties, portable number and Aadhaar card subtleties of the vehicle proprietor. In the exploration
article standardized identification recognition for ETC frameworks has been proposed. The present work accentuates the
execution of the proposed scanner tag location strategy so as to get better efficiency. The execution of the method
represented in this paper shows comprised results. A significant improvement is portrayed during the time spent
standardized identification discovery, over the conventional approach.
Image-Based Barcode Detection and Recognition to Assist Visually Impaired Persons.
I. Wendy P. Fernandez, Yang Xian, Student Member, IEEE, and Yingli Tian, Senior Member, IEEE The
7th Annual IEEE International Conference on Cyber Technology in Automation, Control and Intelligent Systems
July 31-August 4, 2017, Hawaii, USA
A great many individuals overall experience the ill effects of visual debilitation or visual deficiency. One of the real
confinements brought about by the diminishing in visual capacities is the troubles experienced during shopping. In this
paper, we propose a standardized identification location and acknowledgment strategy for a shopping right hand. The
proposed technique is fit for removing the basic item data (e.g., fixings, source, calories, and so on.) from the identified
standardized tag locale. Proposals of human framework association are given to improve the client experience. The
shopping associate encourages the procedure of shopping for food and expands the independency of outwardly debilitated
people. To accomplish a constant location and acknowledgment, we construct the identification structure upon a state-
ofthe-craftsmanship parallel fragment indicator that takes points of interest of the property that standardized
identification comprises of a gathering of parallel lines. Standardized identification acknowledgment is practiced using a
hearty telephone application that gives the point by point data of the item progressively. As exhibited by the test results,
the proposed strategy is compelling, vigorous, and demonstrates incredible potential to be ventured into an assistive
framework for outwardly weakened people.
Visual hindrance is the decrease of vision not fixable with standard glasses, contact focalpoint, drug or medical procedure .
285 million individuals are assessed to be outwardly weakened worldwide and 39 million are visually impaired . Visual
disability, particularlyvisual deficiency, diminishes the capacity of individuals to perform numerous day by dayexercises.
One of the most troublesome exercises is shopping for food since it normally depends on human visual framework to
remove the helpful item data. Scanner tags are institutionalized item identifiers used to build data the executives
proficiency. They are the primary item labeling strategy at supermarkets. Item acknowledgment by outwardly debilitated
clients is frequently performed dependent on filtering gadgets, magnifiers to peruse the marks or located shopping
colleagues. The assets are costly and in some casesinaccessible particularly when it requires the nearness of someone else.
A few telephone applications offer standardized identification perusers. In any case, these applications depend on the
clients to precisely restrict the standardized tag areas, and accordingly, areas yet not appropriate and advantageous for
visually impaired users. The proposed technique takes an item picture caught by the client as the info. It plans to identify
and perceive the standardized identification to remove the item data. The info picture is first bolstered to the line
fragment recognition calculation and a short time later, each line section is assessed to quantify its probability having a
place with the standardized tag area. A standardized identification competitor area is produced after the scoring capacity.
This up-and-comer area is then edited, spared, and filled in as the contribution to a scanner tag acknowledgment
application (i.e., ShopWell) to extricate the essential item data for the client. In situations where the scanner tag isn't found
during the discovery procedure, or the acknowledgment application neglects to separate the item data, basic flagging (e.g.,
vibrations or signal tone) could be used as criticism to teach the client to retake the picture. In this paper, we have
introduced a structure that takes an
information picture (or different if no standardized identification caught from the outset endeavor) of a shopping item
caught by the outwardly hindered clients and returns the itemized data of this item dependent on the distinguished
scanner tag district. The proposed scanner tag location and acknowledgment strategy is clarified and assessed usinga self-
gathered standardized identification dataset. We additionally incorporate recommendations to help the clients in taking
the photographs in situations where there is no or halfway standardized tag in the caught picture and permitting
compelling data recovery for explicit inquiries. Our future work will concentrate on structure a start to finish framework,
precision improvement, UI study, and framework assessment and refinement with the cooperation of visually impaired
clients.

J. Quick response barcode deblurring via doubly convolutional neural network Haitao Pu & Mingqu Fan &
Jinliang Yang & Jian Lian
Different picture preprocessing applications for 2D (two dimensional) standardized identification include switching the
corruption activities (for example deblurring). A largeportion of the recently proposed deblurring methodologies center
around the development of reasonable deconvolution models, which have demonstrated huge execution at lab level. Be
thatas it may, the model-based picture deblurring arrangements probably won't function admirably in useful situations. To
manage this issue, a convolutional neural system (CNN) was proposed based structure to handle the sans parameter
circumstance for 2D scanner tag deblurring. The proposed arrangement use the profound learning system to cross over
any barrier between customary model-based techniques and necessity of turning around the foggy 2D scanner tag
pictures. Investigations on for all intents and purposes obscured snappy reaction (QR) standardized tag pictures show that
the proposedmethodology accomplishes the prevalent presentation in correlation with best in class model-based picture
deblurring approaches.
QR scanner tags [35] (as appeared in Fig. 1), which were designed by the Toyota auxiliaryDenso Wavein 994 and were
used during manufacturing of vehicles to track them process, area run of the mill category of lattice 2D standardized tags.
QR barcode iscomprised of a printed square pattern of little highly contrasting squares that encodeinformation which can
be examined into a PC framework.
The black and white squares within it can represent numbers from 0 to 9, letters from A to Z, orcharacters in non-Latin
contents, for example, Japanese kanji [13]. There have been various adaptations of QR scanner tags proposed with various
data limit as appeared in Fig. 2. QR standardized tags are connected generally because of a lot of preferences, including
little tag, huge information limit, unwavering quality, and quick examining, it assumes a critical job in an assortment of
down to earth applications, for example modernrobotization and assembling frameworks [30], open transportation [34],
open cleanliness
[16], mobile phone based application [31, 48], and item following [21, 46]. At present, applicationsbased on the QR codes
can acquire sufficient imaging quality with late development of
optical hardware [44].However, there were issues with the readability of QR code as they may still be declined during
imaging strategy because of different issues including out-of-center haze [32] andmotion blur[6].Not only many existing
algorithms can be defeated using these but also new challenges can be created in the QR code perusing and interpreting
process. As it is intended for the contained data to be decoded at rapid [12], QR code is widely abused for item tracing in
industry production line. Therefore, the deblurring of QR code images becomes one of themost major and significant issues
in mechanical computerization for assembling. In the proposed CNN engineering, one deconvolution procedure is
incorporated to opposite the corrupted functional QR standardized identification picture. In this manner, we initially
talk about the obscuring model and deconvolution activity utilized in our strategy before we present the proposed
technique. Deconvolution activity in our strategy, and so on. In this paper, we propose a novel CNN answer for handle QR
scanner tag picture deblurring task. To our best learning, this is presumably the primary utilization of CNN to take care of
the issue of QR scanner tag picture deblurring. So as to deal with the smoothness in QR standardized tag picture, we
receive a recently introduced convolutional layer in the proposed CNN engineering. With the preparation of hazy QR
standardized identification pictures deteriorated by various haze pieces, our CNN based strategy can be utilized to switch
both the out-of-center obscured and movement obscured QR scanner tag pictures. To affirm the enhanced visualization of
our technique, we led contrasting investigations between cutting edge deconvolution strategies and our proposed
strategy. Besides, to check the coherence of our strategy, we likewise directed looking at examinations between cutting
edge techniques and our own with Zbar [50]. Test results demonstrate that the CNN technique beats the cutting edge
deconvolution strategy to deblurring QR standardized tag pictures both in people's vision and machine vision
meaningfulness.
Through consolidating all the valuable data of information picture and profound learningmethodology, the proposed CNN
arrangement can significantly improve the presentationof deblurring QR standardized tag picture.
K. Real-Time Barcode Detection and Classification using Deep Learning Daniel Kold Hansen, Kamal
Nasrollahi, Christoffer B. Rasmusenand Thomas B. Moeslund
Standardized tags, in their various structures, can be found on practically any bundles accessible in the market. Identifying
and after that unraveling of standardized identifications have subsequently extraordinary applications. We depict how to

adjust the condition of-theart profound learning-based indicator of You Only Look Once (YOLO) todistinguish scanner tags
in a quick and solid way. The locator is equipped for recognizing both 1D and QR standardized identifications. The finder
accomplishes best in class resultson the benchmark dataset of Muenster BarcodeDB with a location pace of 0.991. The
created framework can likewise find the revolution of both the 1D and QR standardized tags, which gives the chance of
turning the discovery in like manner which is appeared to benefit the interpreting procedure in a positive manner. Both
the recognition and the revolution expectation demonstrates continuous execution.
Keywords: Deep Learning, Barcode detection, Barcode Rotation
Scanner tags are a coordinated piece of the present reality and are utilized in a widerange of settings extending from the
neighborhood general store to the utilization in publicizing. Standardized identifications can be part into two diverse
fundamental categories,1Dand 2D scanner tags. The best realized 1D scanner tag types are presumably the EAN an UPC
type which is primarily utilized for naming shopper items at the neighborhood market. An extremely known and well
known 2D scanner tag is the QR standardized tag. The QR standardized identification is for instance utilized in showcasing
where it goes about as a connection between the printed and advanced media, by diverting individuals to extra data,
rivalries, internet based life destinations, and so on. To decipher standardized identifications, a few arrangements exist
going from laser scannersto camera based gadgets. Customary arrangements, for example, the laser scanner don't give the
opportunity of decoding 2D barcodes, to do that camera based scanners are required.
A mainstream camera based scanner is the cell phone which enables the client to filter forall intents and purposes any kind
of standardized identification. The cell phone does, be that as it may, requires a specific measure of direction from the
client, and are generally just equipped for unraveling one standardized tag at the time. To streamline this procedure, it
could be attractive to find standardized tags in a picture and in this manner have the option to interpret different scanner
tags at the time and require less direction from a client. Profound learning has been fruitful in different territories beating
different strategies.
In the field of standardized identification localisation, the main scanner tag locator arrangement known to the creator,
utilizing profound learning is Zamberletti2013, where it is utilized to investigate a Hough Space to find potential bars. We
might want to explore whether the utilization of profound learning can benefit the situating of standardized tags and
accomplish best in class results. We will utilize the profound learning object identification calculation You Only Look Once
(YOLO) (Redmon and Farhadi, 2016) for finding the standardized tags. We will attempt to prepare the system to have the
option to distinguish 1D scanner tags (UPC and EAN) and the QR standardized tags. We will utilize the YOLO system
dependent on Darknet19 with the information size of 416x416. The following characteristic advance in the wake of finding
a standardized tag is interpret it.
Through some little scale test, we discovered that turning the 1D standardized identifications to such an extent that the
bars are vertical and pivoting QR scanner tags sothe sides of the little squares line up with the x and y axis can benefit the
exhibition of the unraveling. For 1D scanner tags, there is a speedup in time and a higher deciphering rate, though for the
QR standardized identifications the unraveling will take longer, yet the translating achievement rate is higher. To find the
measure of turn required a relapse system is utilized to foresee a pivot an incentive somewhere in the range of 0 and 1.
The worth will be mapped to a point going from 0 to 180 for 1D and 45 and 135 for QR barcodes.
Atfig.2themethodonhowtheangleis estimated is appeared. The relapse system depends on the Darknet19 classification
network1 where the softmax layer is evacuated, and the quantity of filters in the last convolutional layer is set to one.
Moreover, three distinctive actuation capacities are attempted in the last convolutional also, Leaky ReLU, Logistic and
ReLU.
We told the best way to utilize profound learning to identify scanner tags in a picture. Theidentifier has demonstrated to
be vigorous with condition of the outcomes on the Muenster Barode DB. Furthermore, it has been demonstrated that we
can distinguish both1D and QR scanner tags with a similar system and extra standardized tag types can undoubtedly be
included. Other than preparing a system for scanner tag detection, a arrange ready to foresee the point of pivot of
standardized identifications. The system for anticipating the point is a relapse system dependent on the Darknet19
engineering, which was prepared and tried for both 1D and QR barcodes. The trial of how the point expectation can benefit
the disentangling of the scanner tags demonstrated that the forecasts gave a raise in the unraveling achievement rate for
every one of the tests.
Besides, the ZXing 1D standardized identification translating gave a speedup in theunraveling time.

L. Studying the Effect of Paralleling Settings on theFunctioning of a Barcode Recognition App Ventsov
N.N., Podkolzina L.A. 2017 International Conference on Industrial Engineering, Applications and Manufacturing
(ICIEAM)
The paper proposes a parallel video stream preparing strategy for a direct standardized identification acknowledgment.
The scientists have reenacted the framework for different memory utilization modes. Various investigations have been
completed to recognize the shortcomings and characteristics of the framework. In light of these information, upgrades
were made to tackle the issue of an increasingly proficient distinguishing of the articles in a video stream. Hypothetical
investigations were completed based on the parallel processing hypothesis, vector variable based math and the example
acknowledgment hypothesis. The methodology empowers a progressively proficient utilization of processorabilities.
Keywords— computer vision; pattern recognition; 1D barcodes; barcode detection; multithreading; video stream
processing; parallel computing
Problem Statement+Methodology+Expected Output
This paper abides upon the issue of recognizing straight scanner tags based on the Houghchange. The examination article
is a lot of video records containing straight standardizedtags. A video stream preparing framework ought to be built up
that utilizes parallel registering. At that point the working of the framework ought to be examined dependent on the yield
acquired, and its exhibition ought to be enhanced. Programming and equipment restrictions ought to be considered.
Hypothesis Single-strung frameworks are not quite the same as parallel frameworks; this distinction lies in the utilization
of multithreading in the most requesting tasks. There are three parallel figuring plans: balanced strings, pipelining, and
consolidated methodology [17-23]. To tackle this issue,we'll utilize balanced strings for a similar activity (which includes
one video stream outline). Parallel calculation speeding up is viewed as the proportion of the run-time of the best
sequential calculation to the run-time of its parallel partner. The framework calculation comprises of a few phases. To
begin with, information from the video streamare stacked into the info cushion. In the subsequent stage, the video stream
information are prepared. Information are recovered from the info cradle, handled and put away in the yield cushion.
Yield video stream, or video record, is produced in the third arranges. Every one of these moves are made at the same
time. Activities of the main stage and thethird stage are taken care of by independently running strings. The subsequent
stage is separated into different simultaneous strings. It has been considered that races happen in multithreaded video
preparing because of the strings being nonconcurrent and between autonomous, which eventually issue the casing
arrangement. This is the reason each casing is allocated a remarkable ID to create the yield video stream appropriately
This examination delivered a data framework for multithreaded video information handling to recognize straight
standardized tags utilizing PC vision innovation based on the Hough change.
Various tests has been completed to distinguish the shortcomings and idiosyncrasies of the framework. It has been
discovered the multithreading component increments when changing from yield video gushing to creating a yield
document depictingthe distinguished item. The strategy proposed was streamlined dependent on this outcomes for an all
out 10% expansion in framework execution.
M. Vision bаsed Distаnce Meаsurement System using Two dimensionаl Bаrcode for Mobile Robot Jong Hwаn
Beck, Sаng Hoon Kim
Productivity problem in the future might be solved by using robots based on vision technology. Dueto their characteristics
of using the camera, they will be very useful in various sectors. Landmark detection technology using cameras is utilized
here becаuse it is eаsy to meаsure the position. In this paper, a system for measuring distance has been proposed that
works by detecting two dimensionаl bаrcode which cаn be used in the building of unmanned ground vehicles. In
experiments, the position ofthe cаmerа is expressed аs (x, y) by using 2D barcode recognition in a 4 by 4 grid.
Keywords—mobile robot, locаlizаtion, bаrcode detection, lаndmаrk, QR code Problem
Stаtement+Methodology+Expected Output
Studies have suggested that later in this century, birth rates will decline and technology will become cheaper which would
result in the economy and production depending on robots. Recently, smart andintelligent robots have been developed as
the technology keeps advancing and the processing power of computers grows day by day. Intelligent robot аre expected
to be deployed in industriаl as well as human household spaces as with advancing technology, it also becomes cheaper and
more people canafford then. [1]. Researchers and experts believe that to be able to function in home environments, high
level precision is required. For this precision to be achieved these robots need to have a good sense of where they are with

respect to the other objects in the household as well as where they are at all times. Various methods have been introduced
that use a combination of numerous sensors to achieve this. Self-positioning or localization is achieved using two things,
incremental method and absolute method. Incremental method would be the sensors to give the robot an immediate
mapping of surroundings while absolute method would be something like GPS. Mаp mаtching, аnd lаndmаrk recognition
аnd so on will be needed by this technology. Image based landmark recognition would bevery important and useful here.
While а bаrcode is blаck аnd white bаrs of different thicknesses that a computer can read, in vision-bаsed systems,
bаrcodes аre used аs representаtive objects. Here, a distance measurement system is proposed that utilizes barcode
recognition. In pаrticulаr, we used two-dimensionаl bаrcodesfor this method. The purpose here is to develop a system that
is robust and can overcome the drawbacks that come with the existing technology and helps simplify the process of
building these indoor robots. A vision based distancemeasuring system is used. The QR code cаn contаin more informаtion
compаred to the 1D bаrcode, so it is helpful for our purposes. For future work, improvements in the algorithms would be
the aim as that would makethe distance measuring process that much faster.
2.Methodology
2.1 Understanding Barcode Recognition
Figure 2: Project Block Diagram
The undertaking displayed the approach as appeared in Figure 2 above. Right off the bat, picture that contains the
standardized identification data is gained either by utilizing a camera or chose picture from any record from PC.
The shading picture contains in truth the full usable data. The picture is then being changed overto grayscale design. The
picture is changed into a pre-prepared grayscale picture for lessening commotion and upgrading the picture differentiate
among bars and spaces.
A scanner tag direction is recognized that is parallel to bars in the pre-handled picture. An anticipated succession is
created by anticipating pixels of the pre-handled picture along the scanner tag direction. Since the gained picture will
commonly contain a territory bigger than the standardized identification, it is at first required to edit the scanner tag
region as the remainder of the picture is superfluous. Since the standardized tag area is required, the outskirts of the
scanner tag line directions must be resolved. An edge identification calculation was utilized to decide the fringes. The
binarized arrangement is created by applying an edge to distinguished pinnacles and valleys of the anticipated succession.
A standardized tag encoding grouping at that point is created from the binarized arrangement, in which each scanner tag
bar module and each scanner tag space module are spoken to by a solitary separate piece in the standardized tag encoding
succession. The standardized tag encoded data is extricated from the scanner tag encoding arrangement

2.2 Pre-processing Techniques
Preprocessing methods are required on shading, dark level or twofold report pictures containing content or potentially
designs. In character acknowledgment frameworks the majority of the applications utilize dark or parallel pictures since
handling shading pictures is computationally high. Such pictures may likewise contain non-uniform foundation as well as
watermarks making it hard to separate the report content from the picture without playing out some sort of
preprocessing, in this way; the ideal come about because of preprocessing is a twofold picture containing content as it
were. Hence, to accomplish this, few stages are required, first, some picture improvement strategies to expel commotion
or right the difference in the picture, second, thresholding to evacuate the foundation containing any scenes, watermarks
and additionally clamor, third, page division to isolate illustrations from content, fourth, character division to separate
characters from one another and, at last, morphological handling to upgrade the characters in situations where
thresholding as well as other preprocessing systems disintegrated portions of the characters or added pixels to them. The
above strategies present few of those which might be utilized in character acknowledgment frameworks and in certain
applications; few or a portion of these procedures orothers might be utilized at various phases of the OCR framework. The
remainder of the part will show a portion of the strategies utilized during the preprocessing phase of a character
acknowledgment framework.
2.2.1 Image Enhancement Technique
Image enhancement improves the quality of images for human perception by removing noise, reducing blurring,
increasing contrast and providing more detail. This section will provide some of the techniques used in image
enhancement.
2.2.1.1 Spatial Image Filtering operations
Pictures caught regularly might be impacted by commotion; be that as it may, the subsequent picturesmay not give wanted
pictures to investigation. What's more, in pictures with worthy quality, certain locales may should be underscored or
featured. Spatial preparing is grouped into point handling and cover handling. Point preparing includes the change of
individual pixels freely of different pixels in the image. These basic activities are ordinarily used to address for
imperfections in picture obtaining equipment, for instance to make up for under/over uncovered pictures. Then again, in
cover handling,the pixel with its neighborhood of pixels in a square or circle veil are engaged with producing the pixel at
(x, y) facilitates in the improved picture.
2.2.1.1.1 Contrast Stretching
The degree of difference in a picture may shift because of poor light or ill-advised setting in the obtaining sensor gadget.
Along these lines, there is a need to control the complexity of a picture so as to make up for troubles in picture securing.
The thought is to change the dynamic scope of the dim levels in the picture. A system that could work for this situation is
called direct mapping, condition (2), to extend the pixel estimations of a low-differentiate picture or high-differentiate
picture by broadeningthe dynamic range over the entire picture range from 0 – (L-1).
(2)
where O1 relates to 0 and O2 compares to the quantity of wanted levels which is (L-1 = 255). I1 and I2 givethe base and
greatest estimations of the information dim level range. The least complex type of preparing is to change the brilliance of a
picture by including a predisposition esteem, b, to all the pixel estimations of a picture; where b > 0 would expand the
splendor of a picture and b < 0 would obscurethe picture. Additionally, an increase factor, a, might be utilized rather than an
inclination, where the resultof a with the information pixel esteems alter the splendour of the yield picture.
Estimations of 0 < a < 1 will create a darker picture and estimations of a > 1 will deliver a more brilliant picture. Joining
both predisposition and increase produces condition (3).

(3)
For this situation, we have to indicate both the increase and predisposition esteems, yet in common sense it might be hard
to do as such; accordingly, the arrangement is map the info picture extend (I1,I2) to the yield picture run (O1, O2 ) where
O1 compares to 0 and O2 relates to the quantity of wanted levels, subsequently direct mapping characterized in condition
(2).
2.2.1.1.2 Log Transformation
(4)
where c is a constant and it is assumed that r ≥ 0. This transformation maps a narrow rangeof low grey-level values in the
input image into a wider range of output levels and vice versa.
2.2.1.1.3 Power Transformation
(5)
where c and γ are positive constants and is an offset which is usually ignored since it is due to display calibration.
Therefore; s=c(r+e)gamma , where values of 0 < γ < 1 map a narrow range of dark input values into a wider range of
output values, with the opposite being true for values of γ greater than
1. This demonstrates the power-law changes are significantly more flexible in such application thanthe log change. Be that
as it may, the log capacity has the significant trademark that it packs the dynamic scope of pictures with huge varieties in
pixel esteems. Because of the assortment of gadgets utilized for picture catch, printing, what's more, show react as per the
power law type, gamma, (γ), this factor should be amended, consequently control law reaction wonders or gamma remedy
which is given by (5).
2.2.1.1.4 Sharpening (High-pass) filter
A honing channel is utilized to underline the fine subtleties of a picture (i.e., gives the inverse impact of smoothing). The
purposes of high differentiation can be identified by figuring force contrasts in nearby picture districts. The loads of the
veil are both positive andnegative. At the point when the cover is over a region of consistent or gradually changing dim
level, the aftereffect of convolution will be near zero. At the point when dark level is changing quickly inside the area, the
consequence of convolution will be an enormous number. Commonly, such focuses structure the fringe between various
articles or scene parts(for example edge). A case of a honing channel is the Laplacian channel which is characterized in
condition (6) beneath.
(6)
This execution can be applied at all focuses (x,y) in a picture by convolving the picturewith the accompanying spatial veil
Fig. 4(a) with an elective meaning of the computerized second subsidiaries which considers the corner to corner
components andcan be executed by the cover in Fig. 4(b).

Figure 4. Laplacian Filter
The Laplacian channel is a subordinate administrator which hones the picture, howeverdrives consistent regions to zero;
in this manner, including the first picture back reestablishes the dark level tonality, condition (7).
(7)
Where, f(x,y) is the information picture, g(x,y) is the yield picture and c is 1 if the insidecoefficient of the cover is sure, or -
1 in the event that it is negative
2.2.1.1.5 Skew Correction
Because of the probability of turn of the information picture and the affectability of many report picture examination
techniques to revolution of the picture, archive slant ought to be rectified. Slant location systems can be generally
characterized into the accompanying gatherings: investigation of projection profile, Hough change, associated parts,
grouping, and Correlation between lines procedures. The review by Hull and Taylor, examined twenty-five unique
strategies for record picture slant identification. The techniques incorporate methodologies dependent on Hough
Transform investigation, projection profile, include point dissemination and direction delicate component examination.
The study presumed that the majority of the systems detailed a scope of up to 0.1 degrees exactness, confirming a solid
requirement for further work here to help demonstrate the qualities and shortcomings ofindividual calculations (Hull and
Taylor, 1998). What's more, there are new procedures developing for explicit applications, for example, the technique for
Al-Shatnawi and Omar which depends on the focal point of gravity for managing Arabic record pictures (Al- Shatnawi and
Omar, 2009). Along these lines, the decision of utilizing a slant discovery/revision system relies upon the application and
the kind of pictures utilized.
2.2.1.1.6 Noise Removal
Where
And
(9)
(8)

(10)
2.2.1.1.7 Thinning and Skeletonizing
Skeletonisation is a process for reducing foreground regions in a binary image to a skeletal remnant that largely preserves
the extent and connectivity of the original region while removing most of the original foreground pixels. It is clear to
imagine that the skeleton is asthe loci of centres of bi-tangent circles that fit entirely within the foreground region being
considered. There are two fundamental systems for delivering the skeleton of an article: essential diminishing and average
hub changes. Diminishing is a morphological activity that is utilized to expel chosen closer view pixels from parallel
pictures, to some degree like disintegration or opening. Diminishing is an information decrease process that dissolves an
article until it is one-pixel wide, creating a skeleton of the item making it simpler to perceive articles, for example,
characters. Fig. 5 shows how diminishing the character E creates the thin state of the character. Diminishing is ordinarily
just applied to paired pictures, and delivers another twofold picture as yield. Diminishing dissolves an article again and
again (without breaking it) until it is one-pixel wide. Then again, the average hub change finds the focuses in an article that
structure lines down its inside.
Figure 5: (a) Original Image (b)Medial Axis Transform (c)Outline (d) Thinning
3. Result
3.1 Implementation
The code was done in python with the help if the OpenCV library, we could do most of this.
Code
import numpy as np
import matplotlib.pyplot as pltimport sqlite3
from pyzbar import pyzbarimport argparse
import cv2
##ap = argparse.ArgumentParser()
##ap.add_argument("-i", "--image", required=True,help="path to input image")##args = vars(ap.parse_args())
# load the input image
##image = cv2.imread("qr code.png") cap = cv2.VideoCapture(0)

and applying that kernel laplacian_kernel=np.array([[0,1,0],[1,-4,1],[0,1,1]]) def apply_lap(frame):
lap=cv2.filter2D(frame,-1,laplacian_kernel)return lap
#laplaciam kernel 2 and applying that laplacian_kernel2=np.array([[0,1,0],[1,-8,1],[0,1,1]]) def apply_lap2(frame):
lap2=cv2.filter2D(frame,-1,laplacian_kernel2)return lap2
def apply_invert(frame):
return cv2.bitwise_not(frame)def darktobright(frame):
retval, threshold = cv2.threshold(frame, 12, 255, cv2.THRESH_BINARY)return retval,threshold
def histeq(frame): frame=cv2.cvtColor(frame,cv2.COLOR_BGR2YUV) frame[:,:,0]=cv2.equalizeHist(frame[:,:,0])
equ=cv2.cvtColor(frame,cv2.COLOR_YUV2BGR) return equ
def some(): while True:
ret, frame = cap.read()
#gray= cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)#out.write(frame)
invert = apply_invert(frame) retval,threshold=darktobright(frame) lap=apply_lap(frame) lap2=apply_lap2(frame)
equ=histeq(frame) cv2.imshow('equ',equ) cv2.imshow('lap1',lap)
cv2.imshow('lap2',lap2) cv2.imshow('dtb',threshold)cv2.imshow('invert',invert)
cv2.imshow('frame',frame)#cv2.imshow('gray',gray)
if cv2.waitKey(1) & 0xFF == ord('q'):break
# find the barcodes in the image and decode each of the barcodesbarcodes = pyzbar.decode(frame)
#to output the file
#fourcc = cv2.VideoWriter_fourcc(*'XVID') #out=cv2.VideoWriter('output.avi', fourcc, 20.0, (640,480)) #laplcaian kernel
if barcodes:
break
#cap.release() #out.release() cv2.destroyAllWindows()
# loop over the detected barcodesfor barcode in barcodes:
# the barcode data is a bytes object so if we want to draw it on# our output image we need to convert it to a string first
barcodeData = barcode.data.decode("utf-8")
barcodeType = barcode.type
# print the barcode type and data to the terminal
print("[INFO] Found {} barcode: {}".format(barcodeType, barcodeData))print ("For More Information type y/n")
a=input()
if (a=="y"):
conn=sqlite3.connect('test.sqlite') cur=conn.cursor() data=barcodeData.split('n') idn=data[0].split(":")

#print(idn)
cur.execute('SELECT field2 FROM test1 WHERE ID=( ? )',(idn[1],) )row=cur.fetchone()
for line in row:
print (line)
some() while True:
print("Read more")n=input()
if (n=='y'):some()
else:
cap.release()break
3.2 Output
This program scans the barcode and then once is scans the windows get closed and then it use’s that information from that
barcode and then compares it to the data in the database and the displays us all the information that we need to know
about that product. After that it alsoasks us if we want to scan more barcodes and continue.
idn[1]=idn[1].lstrip()
The windows going form left to right and top to bottom, we have Inverted, Normal frame, Laplacian kernel 2, Histogram
equalizer, Threshold, Laplacian kernel 1. This has improvedthe QRcode scanning in even unfavorable conditions.
Once the QRCode is scanned, it is read from the starting and the value of ID in the first line istaken and the DB
SQLCipher is used to take the ID and then get the desired ID’s information out there.

The data which is shown is then taken and printed, and if we want more information, “y” keyis used and taken as an input
to get more information about it.
3.3 Conclusion
With all these enhancements, we can try to implement them in improving and the barcode and QRCode scanner to work it
less optimal conditions. The limitation that we can face are the ones where we see the problem of size of the barcode or
QRCode, after a certain size, It is not recognizable to the software of in dark conditions. And with the possible use of
Machine Learning Techniques we can try to use the technique we must in order to read the code without wasting time or
computing power in hand. This gives us the possibility of usingthese implementation on cheap and accessible hardware
and also this will also allow to put barcode on objects much smaller than what we could before, such as computer
standoffs and screws, etc.
4. References
[1] 2018 15th International Conference on Ubiquitous Robots (UR) Hawaii ConventionCenter, Hawai'i, USA, June 27-30,
2018. Hyeon Cho, Dongyi Kim, Junho Park, KyungshikRoh, Wonjun Hwang
[2] Jie Chen, Xianjie Pu, Hengyu Guo, Qian Tang, Li Feng, Xue Wang, Chenguo Hu, 11November 2017
[3] Proceedings of the World Congress on Engineering 2017 Vol I WCE 2017, July 5-7,2017, London, U.K
[4] Automatic Barcode Extraction for Efficient Large-Scale Inventory Management
[5] Autonomous Recognition System for Barcode Detection in Complex Scenes

[9] Electronic Toll Collection System Using Barcode Technology
[10] Image-Based Barcode Detection and Recognition to Assist Visually Impaired Persons
[11] Quick response barcode deblurring via doubly convolutional neural network
[12] Real-Time Barcode Detection and Classification using Deep Learning
[13] Studying the Effect of Paralleling Settings on the Functioning of a BarcodeRecognition App
[14] Vision based Distance Measurement System using Two-dimensional Barcode forMobile Robot
[15] 2017 International Conference on Industrial Engineering, Applications (ICIEAM)
[6] Barcode Character Defect Detection Method Based on Tesseract-OCR
[7] Rahmah Al Sheikh, Raghad Al-Assami, Maryam Albahr, Muntaha Al Suhaibani, Mutasem k. Alsmadi, Muneerah
Alshabanah, DaniahAlrajhi, Ibrahim Al-Marashdeh, Sanaa Alsmadiand Hayam Abouelmagd and Mohammed Tayfour
[8] Rama Rani, Gaurav Deep. 4th IEEE International Conference on Signal Processing, Computing and
Control(ISPCC 2k17), sep21-23, 2017, Solan, India.

Limitations of Barcode and QR code Scanner

More Related Content

Similar to Limitations of Barcode and QR code Scanner (20)

More from IRJET Journal (20)

Recently uploaded (20)

Limitations of Barcode and QR code Scanner