Humans have an outstanding ability to recognise, classify and discriminate
objects with extreme ease. For example, if a person was in a large classroom
and was asked to find the light switch it would not take more than a second or
two. Even if the light switch was located in a different place than the human
expected or it was shaped differently than the human expected it would
not be difficult to find the switch. Humans also don’t need to see hundreds of
exemplars in order to identify similar objects. For example, a human needs to
see only a few dogs and then he is able to recognise dogs even from species that
he has not seen before. This recognition ability also holds true for animals, to
a greater or lesser extent. A spider has no problem recognising a fly. Even a
baby spider can do that. At this level we are talking about a few hundred to a
thousand processing elements or neurons. Nevertheless the biological systems
seem to do their job very well.
Computers, on the other hand, have a very difficult time with these tasks.
Machines need a large amount of memory and significant speed to even come
close to the processing time of a human. Furthermore, the software for such
simple general tasks does not exist. There are special problems where the
machine can perform specific functions well, but the machines do not perform
general image processing and recognition tasks.
In the early days of electronic image processing, many thought that a
single algorithm could be found to perform recognition. The most popular of
these is Fourier processing. It, as well as many of its successors, has fallen
short of emulating human vision. It has become obvious that the human uses
many elegantly structured processes to achieve its image processing goals,
and we are beginning to understand only a few of these.