Global Positioning System Reference
In-Depth Information
7.4.2
Inside-Out Positioning
Another facet of the digital camera revolution is the spread of personal cameras
that are now being embedded into other appliances (i.e., mobile phones,
binoculars, and laptop PCs). In these cases the positioning is inside-out (i.e., the
camera looks out at a scene that will vary as the camera is moved). By comparing
what can be seen with what it might be possible to see, it is possible to determine
both location and orientation. For example, Cipolla and Robertson [4] have
developed a system where a known scene captured previously by a camera can be
parameterized (or turned into metadata by encoding the shapes within the scene).
This parametric signature of the scene can then be tagged with other information
such as location and a description. It is then stored within a library of common
scenes (e.g., famous buildings, landscapes, and objects). When a user takes a
photograph, a service can be offered to compare its signature with those in the
library. If a match is found, location can be found. From the shape and angle
differences it is also possible to gain information about relative distance and
orientation between the user and the main subject in the scene.
7.4.3
Ubiquitous Digital Cameras
Before looking at some applications and techniques, it is important to know why
digital cameras are becoming so ubiquitous and how camera and processing
technology is developing. Since the first transistors were made in the early 1950s,
it was evident that they were all light-sensitive and measures were taken to keep
light away from the active regions by encapsulation. On the other hand, the light-
sensitive phenomenon is useful it its own right and a whole range of photo
transistors, diodes, and more exotic optoelectronic components have followed.
A computer memory chip is made up of a grid of millions of adjacent
transistors. If it is exposed to light, it becomes an electronic version of an eye's
retina and all that is needed to turn it into a camera is a simple lens to focus an
image. Since memory chips have been falling in price and growing in capability in
line with computers in general, it is not surprising that digital cameras have also
become very cheap. There are many specialized camera chips: the most basic are
CMOS arrays and those of higher performance are charge-coupled devices
(CCDs). The continuation of general improvement in silicon chip technology,
which includes the cost reduction per bit, an increase of component density, and
an increase in the switching bandwidths of the transistors, is known as Moore's
law. It is a holistic relationship spotted by one of Intel's founders, Gordon Moore,
in the 1960s. It has been true for 50 years and is predicted to continue to hold for
at least another decade. We can expect, therefore, that the spread of basic
components like simple cameras will continue to the point where anywhere there
is activity, there will usually be cameras and most of them will be networked.
Whatever other positioning technology may be available for an application,
there is also likely to be cameras close by. Taking their images and using machine
