Computer vision describes the ability of computers or subsystems to identify
objects, scenes and activities in images. To this end, technologies are used
with the help of which the complex image analysis tasks are divided among
as small sub-tasks as possible and then computed. These techniques are
applied to recognise individual edges, lines and textures of objects in one.
Classification, machine learning and other processes, for example, are used
to determine whether the features identified in an image probably represent
an object already known to the system.
Computer vision has multifaceted applications, among them the anal-
ysis of medical imaging to improve prognoses, diagnoses and treatment
of diseases or facial recognition on Facebook, which ensures that users are
automatically recognised by algorithms and are suggested for tags. Such sys-
tems are already used for security and surveillance purposes for the identi-
fication of suspects. In addition, e-commerce companies such as Amazon
are working on systems with which specific products can be identified on
images and subsequently be purchased directly online. Whilst researchers in
the field of computer vision are working on the aim of being able to utilise
systems independent of the environment, with machine vision, sensors are
used with the help of which relevant information can be captured within
restricted environments. This discipline is technically mature to the extent
that it is no longer part of ongoing informatics research, but part of sys-
tem technology today. At the same time, it is less a matter of recognising
the meaning or content of an image but of deriving information relevant for
action.
Comments
Post a Comment