Information Technology Reference
In-Depth Information
The advantages of desktop and handheld scanners are that they can scan all types
of paper documents without additional provisions made to the documents. However,
only one static picture is taken at a given point in time. So, it is not possible to
continuously track changes made to the document. Moreover, all visual contents are
digitized in one single layer. For example, it is therefore a complex task to separate
the underlying printed document from handwritings made on it, such as annotations
and sketches. Hence, scanning is not optimal for interactive applications.
Another approach supports more interactive uses. The contents of documents
are captured by one or several cameras which are mounted above the table or in
front of an interactive wall. While contents can be continuously tracked, camera
capturing provides lower resolutions than most desktop and handheld scanners. We
will discuss this approach in more detail below.
2.1.2 Page Identification and Location Tracking
Many settings that integrate paper with computing require that a physical sheet of
paper can be uniquely identified. For instance, a physical paper card could be used
as a physical token to access a specific digital object. Alternatively, the system could
help the user to find a paper document in the office by indicating the physical lo-
cation of the document. In both cases, the system must identify the document. Two
main approaches can be distinguished: marker-based and content-based identifica-
tion. Marker-based approaches require that the objects that are to be tracked con-
tain a machine-readable tag. This tag can be visible to the human eye, such as a
printed barcode, or invisibly integrated into the object, such as an electronic RFID
tag. Content-based techniques do not interfere with the visible artwork, but result
in lower processing speed and can distinguish between only a smaller number of
objects than marker-based techniques.
Visual Markers
Visual markers (also called fiducials) encode an identifier in an optical machine-
readable representation. A marker is captured by one or several cameras, by a light
sensor or by a laser scanner. The most widespread form of visual markers, contained
on almost any product, is the linear barcode. It encodes a binary sequence by varying
the width of black bars that are arranged in a linear sequence. The EAN 13 coding
scheme [44] is used worldwide to identify products at cashpoints. Figure 2.2 (a)
shows an example of this barcode.
Linear barcodes are not used by many Pen-and-Paper Interfaces. More common
are two-dimensional fiducials, since they allow for storing more data and also for
tracking the location of objects in 3d space. Several types of two-dimensional fidu-
cials can be distinguished:
