25-07-2012, 12:08 PM
2D COLOR BARCODES FOR MOBILE PHONES
2D COLOR BARCODES FOR MOBILE PHONES.pdf (Size: 424.14 KB / Downloads: 151)
Introduction
Barcodes are optical machine-readable representations of data, capable of storing
digital information about the physical object to which they are attached. Due to
their reading speed, accuracy, and functional characteristics, barcodes have become
ubiquitous in many applications, including their usage in department stores and
retail chains to price goods, to track items and to identify customers through membership
cards; in tracking item shipment and movement, such as express mail, rental
cars, airline luggage; in patient identification in hospitals; in document management
systems; in ticketing for sports events, cinemas, theaters and transportation.
Traditional barcodes, referred to as one-dimensional (1D) barcodes, represent
data by varying the widths and spacings of parallel lines. The amount of digital
information stored in 1D barcodes is limited and could be simply increased
by increasing the number of barcode digits or by laying out multiple barcodes.
This approach has many negative effects, however, such as enlarged barcode areas,
∗Work partially supported by the EU under Contract no. FP7-SME-2010-1-262448 (Project
SIGNED). A preliminary version of this paper was presented at the International Multiconference
on Computer Science and Information Technology [Grillo et al.
matrix barcode layout.
more complex reading operations, and increased printing costs. For this reason, the
barcode technology has been deploying geometric patterns (such as squares, dots,
triangles, hexagons) in two dimensions: such barcodes are referred to as bidimensional
(2D) codes. Note that 2D codes increase the data space available by storing
information in two dimensions, whereas 1D codes contains data in one dimension
only. Figure 1 shows examples of 1D and 2D barcodes.
Available 2D codes span from repeating a single 1D barcode over multiple rows
to exploiting bidimensional shapes in order to represent data. Figure 2 illustrates
the evolution of 2D barcode technology. In particular, Figure 2 (a) shows a multiple
barcode layout: the main disadvantage related to this simple 2D layout is the need
of multiple scans in order to get all the information contained in the barcode. Figure
2 (b) illustrates a stacked barcode layout: in this case one single scan is enough to
obtain the stored information but the scanning equipment must be carefully aligned
with the barcode orientation. Finally, in Figure 2 © a matrix barcode layout is
presented: this layout enables to acquire information with one single scan and does
not require the accurate alignment of the scanning equipment.
There are more than 20 types of conventional 2D codes. Figure 3 illustrates
some examples of 2D codes; one of the main differences among those barcodes lies
in the amount of data which can be stored in a single barcode. For example, the
Quick Response (QR) codea is able to store more than 7,000 decimal digits, while
Maxi code is able to store only 138 decimal digits.
Due to the plethora of applications for barcodes, there is an emerging need
for barcodes capable of storing even more information and more character types
in smaller printing space. Furthermore, the wide diffusion of smartphones equipped
with low-end cameras introduces new challenges for barcodes and makes it appealing
to apply this technology to mobile applications. Just to mention only one application,
the International Air Transportation Association (IATA) recently announced a
global standard for global mobile phone check-in using 2D barcodes [IATA (2007)];
this will allow airlines to send boarding passes as 2D barcodes to the customers’
mobile devices, thus eliminating completely the need for paper travel documents.
Both the increasing demand for higher density barcodes and the wide availabila
QR code is a 2D code developed by Denso Wave, a division of Denso Corporation at the time,
and released in 1994 with the primary aim of being easily interpreted by scanner equipment. QR
code is registered trademark of Denso Wave Incorporated in Japan and other countries.
ity of on-board cameras in mobile phones seem to motivate naturally the need for
2D color barcodes. However, the increased data density obtained with the usage
of colors comes at an additional cost. Today, a Print&Scan process is commonly
used for image reproduction and distribution, where images are converted between
printed and digital formats. A rescanned image may look similar to the original, but
it may have been distorted during the Print&Scan process. Indeed, reading color
barcodes poses significant computer vision challenges [Parikh and Jancke (2008);
Bulan et al. (2009); Siong et al. (2008)]. This is due to several factors, and we cite
only few of them in the following. First, the color balance may be drastically different
in different code readers. Second, the images containing codes may be taken by unexperienced
users, and thus the location of the barcode in the image, its orientation,
its slope, etc. can be mostly unconstrained. Furthermore, possible transformations
in the prospective can distort the geometry of the barcode. Last but not least, the
light conditions under which the images are taken can vary dramatically.