Automation is one of the most important steps in the optimization of production processes towards Industry 4.0. Machine vision is a suitable technology for all sectors and industries, as it can improve quality and increase the efficiency of production lines. Image processing offers several advantages for various applications: It opens new possibilities for efficient quality and production control as well as for warehouse and production logistics.

Generally, an image-processing system consists of a camera, lens, lights and cables, along with a processing unit and control electronics for synchronization and interaction with the production line. The successful development of such systems typically involves a multi-step approach comprised of five project phases: rough specification, feasibility study, detailed specification, system implementation and system integration. This White Paper introduces the three predominant concepts of imageprocessing systems and will help production managers to choose the approach best suited to particular projects and conditions. The software in "classic" image-processing systems must be individually programmed by experienced development teams to provide maximum flexibility in the selection of hardware and software libraries.

Development costs tend to be very high and are often difficult to estimate accurately in advance. The classic approach is recommended if an application requires very specific algorithms or if it is expected to utilize all available computational power. The cost-effectiveness of this approach increases as the number of deployed systems grows. A more recent approach combines traditional image-acquisition and processing hardware with image-analysis methods from the field of artificial intelligence-the so-called self-learning algorithms. This approach accelerates the development of test systems especially for natural materials and other objects, which have high variance that make analysiings with classical approaches difficult. The most innovative approach, based on Smart Cameras, leads the way towards a new generation of image-processing technology. These complete solutions integrate the full set of image-processing components and a software package with an easy-to-use graphical programming interface. With this concept, less-experienced developers and production engineers are able to implement complex systems quickly at low risk. The effort required for development, testing and documentation is typically only a fraction of that under the "classical" approach.

Many factors and criteria influence the selection of an approach that is most appropriate for a specific challenge in manufacturing automation. A concluding table in Section 6 serves as a checklist and individual decision aid, and lists relevant questions and answers.

Do you want to know more? Just download the free Whitepaper. You just need to register on this page.