A camera mounted to a robot is not a new concept by any means and the idea was first pondered in the 1950s. When a robot must be controlled by a human operator who is not local to the machine, cameras must be used to relay information to the operator. Remote controlling of machines has changed dramatically since the onset of the Internet, as operators can now control automated machines from locations hundreds of miles apart. This is particularly useful for controlling machines in environments not reachable by humans such as areas miles below the earth’s surface or deep in outer-space.
Operators responding to a visual display by manually controlling a machine is relatively straight forward compared to robots interpreting what is on a visual display and responding to this information. This is still deemed a daunting, if not impossible task by many engineers working in this field but biometrics has been used to resolve some of these shortcomings and vision systems are now widely used across many industries. Facial recognition technology is used in security systems as an alternative to electronic keys to provide a more secure access system. Facial structure, skin tone and eye colour can all be interpreted by these systems and access decisions can be made by cross checking this information with locally stored data. Although not perfect, vision systems are getting more competent at recognising human faces, tracking objects, deducing colour and interpreting three dimensional modelling.
There are still some deficiencies with these systems and we are a long way off relying on information outputted from these machines. In many applications the camera must be static to interpret accurately what is going on around it and this makes its use very impractical when attached to mobile systems. Amongst other issues; they are unable to differentiate between sexes, objects in motion remain undetectable and the construct of objects is also beyond their functionality. The potential of this technology is great and at present we are only in the infancy of what we will one day be able to achieve with vision systems.