Chapter 17. Practical aspects and applications.
Having discussed a wide variety of facets of vision systems and demonstrated how, in most cases,
it is possible to provide computer algorithms capable of extracting relevant information both
efficiently and simply, it is considered worth spending a little time assessing some of the practical
aspects and applications of computer vision. Under this heading we have three major topics to
explore. Firstly, before any computer or machine vision system can be of much practical use, we
must consider the size of equipment and the time taken for computer processing. In the past much
of the sophisticated image processing for computer vision has required large computers and/or a
long time, in order to extract relevant information from a single image or a short image sequence.
Whilst there are some situations where such size and time penalties are acceptable, in order to
provide versatile computer vision systems it is desirable, even necessary, to be able to carry out near
real time processing on compact and lightweight equipment. Secondly, it is worth considering at
least some of the range of tasks to which computer vision may be applied and trying to relate these
to the wide armoury of techniques available to us. Finally, it is worth considering what, if anything,
we may learn or infer about human (and animal?) vision as a result of the behavioural characteristics
which have emerged during our wide-ranging algorithm development studies.