PATTERN EXPERT legeo: Data Matrix and Dot Matrix Codes
Barcode reader systems are well known to everybody from the shop next door. Modern cash desk systems are coupled to a
barcode scanner which, using the article ID, not only gives information about the price of the article, but can be used to update the
stock database automatically.
There is a set of different types of these well known linear barcodes. Most of them can only be used to encode numerical
characters. In addition, the amount of data encodable in linear barcodes is relatively small.
To overcome these limitations, two-dimensional codes have been developed in the 1970s. Today, Data Matrix Code is the most
popular of these. This code is based on an "International Symbology Specification" by the AIM - Association for Automatic
Identification and Data Capture Technologies - and may be used freely without any license fee. Up to 2335 alphanumerical
characters can be encoded in one Data Matrix symbol. The code makes use of sophisticated error correction algorithms like
ECC200, so that up to 25% of the data may be incorrect. If even more errors occur so that error correction is impossible, the
algorithms are still able to recognize this fact. In this way, the reader will never output a false result but - as the worst possible
case - no result at all.
While the Data Matrix Code's stage of development is very mature, there is still a number of applications which are quite
challenging for current reader systems. Codes embossed on metal are very useful in mechanical engineering. Since embossed
codes cannot be removed (without leaving some trace), they are particularly interesting e.g. for automobile industries. For these
applications, reader systems are needed which can cope with curved surfaces. Also, embossed codes must remain legible
when they become slightly rusted. Dirty surfaces and the industrial environment these reading machines can be exposed to is
another difficulty that must be overcome.
Initially, the Data Matrix code was developed as a printed matrix of square black and white modules. The reader algorithm suggested
by the AIM is adjusted to that classical case. Embossed codes are, in contrast to that, composed of separated circular
depressions. This results in separated spots on the reader system's image of the code ("Dot Matrix"), which cannot be decoded
by the AIM algorithm.
For reading tasks at large distances and particularly for reader systems in mobile applications a common problem is strong spatial
distortion. In this case as well as in the case of codes on curved surfaces the reader system must perform a reconstruction of
the original plane and rectangular matrix shape.
Particularly in the case of shiny metal surfaces, crucial parameters for the reader system are illumination and angle of sight.
Depending on these conditions, the matrix modules may appear as light spots on a dark background, as dark spots on a light
background, or even both in one and the same symbol.
All the difficulties described above are successfully mastered by the
code reader system PATTERN EXPERT legeo.