In the year 1984, I had initiated a special editorial series titled „Optical Systems for Construction, Development and Manufacturing" for a well-known professional journal for construction. Thus I am in the lucky position of having been able to experience and accompany the development of industrially-oriented image processing almost from the very beginning. In my first comment on this topic, I wrote: „Today, the interaction of the technological findings concerning optical systems, electronics, computer science and microprocessor technology permits us to enter into the age of visual communication and of visually automated machine control and -handling. Technological developments have a life of their own and can not be held back."
That is a statement which could be repeated today.

My boldest expectations of that time were, however, exceeded by far through reality. If I had predicted at that time, what is today regarded as a common, everyday application, then I would have even categorized myself as a boundless dreamer. The „minicomputers" used at that time were, after all, in the size of a small table, the cameras were quite voluminous, and the efforts required for programming were complex and intricate.
Nowadays, the efficiency of those systems would only evoke an indulgent smile.
All technological core segments of image processing - optical systems, electronics, computer science and microprocessor technology - have experienced gigantic evolutionary leaps. The interaction of this progress has led to an enormously dynamic development with regard to innovations across the whole bandwidth of industrial image processing.

More than 10 years ago, image processing was exclusively restricted to PC-supported systems. It has now been approximately 10 years, since the first compact and autonomously functioning so-called vision sensors have entered the market. At that time, the heated but technically sound debate about this topic was already to be found in the newest edition of INSPECT. The efficiency of these IP-systems has increased so rapidly within the past 10 years that they have assumed many of the tasks of PC-supported systems. The vocabulary used takes this into account by the fact that the „vision sensors" are now only referred to as vision systems. Hardly five years have passed, since a new performance category of miniaturized systems has established itself, which are now labeled as vision sensors. And today, we are at the verge of another new generation of miniaturized systems in image processing. Parallel to this rapid development, INSPECT has again and again distinguished itself by the fact that it not only had the function of a communication platform for top-notch quality assurance, it also had the ambition to direct the quality of communication towards the requirements of today.

If I should dare to make a forecast regarding the situation in 10 years from now, then I begin to ponder the present explosion regarding performance in all technologies involved in image processing. This will have profound effects in all industrial and non-industrial areas, and in our daily lives as well.
The outcome of extremely effective image processing, in direct combination with a brilliantly conceived control scheme, may already be seen in all walks of life. It is the dream come true for every design engineer involved in image processing. Everything is miniaturized in the greatest extent possible. Extremely fast image capture and transmission of the image information in real time, extremely efficient functioning of algorithms in complex image analysis, fast learning ability in reaction to changed parameters, precise determination of the position and orientation within the changing 3D-space, direct and extremely efficient interaction of image analysis and control functions in the mutually shared processor to direct or command movements. Almost as if everything were on a single chip. And all that is integrated in the heads of insects, such as flies or dragon-flies, measuring only a few millimeters.

Each one of these unique functional modules could, however, also stand synonymously for the dynamic course of development in image processing. This applies for the whole bandwidth of new research projects and developments, with all the fast-track progress they are making. Is the prediction therefore presumptuous or arrogant, that we will be using highly efficient „insect heads" everywhere within the next 10 years, if the framework conditions remain as they are?