Our eyes, as our prior sensory organs, help us to orientate in the world. Through them we can detect the colour, the shape or the movement of the objects.
Machine vision tries to put these features into practice in digital surroundings. Almost every data can be found in the pictures of our cameras from which our eyes produce the above mentioned information for our brains. Starting from this, we can define the picture, as the data source, imbedded in computer surroundings and we can support our machines and tools to do activities depending from them.
As our deeds and thoughts are influenced by the information gained by our eyes, so can machines be taught for the same thing. For this, they should know 2 things: what is it that they can see, and if they recognize it, what to do with it. As if it was an automatic start button, because the visual algorythm itself gives the command for action. And once the recognition is achieved, we can make our devices do almost anything, combining this recognition with other solutions freely and only our imagination can be the limit.
This is a widespread solution in industrial environment, typically in quality control.
With the development of technologies the modern production systems became more and more productive and efficient. During production a lot of emphasis must be done on both quickness of producing them as well as raising the quality of production. The expected quality can be measured on the basis of several parameters. One group of these parameters are the visually perceptible features. If the parameter of a component is visually perceptible, then measuring or perceiving this parameter can also be made automatic, with great likelihood. The components can be classified and identified according to these features, and this can serve as a basis for their recognition. Consequently, we can make simple identifying, recognizing devices, but with the same technology even complex, controlling ones that can examine the tiny details, too.
In civil area it is mainly used in security technology. Let us think of facial recognition-licensing or virtual observation. In automotive developments, too, we can see machine vision more and more often, for instance the road sign recognizing system or the lane-keeping automation are also based on this solution.
This is no question that this technology is getting more and more mature and widespread with the growth of the calculating capacity of the computers, the recognitions and reactions will become more and more complex, so we can build more and more complicated and interesting devices