Image processing has developed into a core technology in many areas of industry. It is difficult to imagine many production processes without the use of image processing, especially at the required quality and low cost. The key drivers of this trend are now the applications themselves: in the automotive industry, with its high hopes of autonomous driving, in robotics, medical applications, not to mention the entertainment industry – just to name the most important areas. There are also many non-industrial fields of application where components with image processing capability are required.
All of this makes the field extremely important from an economic standpoint. In the German image processing industry alone, sales hit Euro1.8 billion in 2014. This was a year-on-year increase of over 10%. Sales growth is expected to be just under 10% again in 2015, bringing industry turnover inches short of the two billion mark.
A study issued in 2013 by the McKinsey Global Institute (Disruptive technologies: Advances that will transform life, business, and the global economy) examined the areas of technology that will have the greatest economic, technological, and societal impact in the future. In many of the areas examined, the study considers image processing as a key enabling technology.
The traditional view of image processing is that it’s used in areas where things have to be precise and quick, or where tiresome tasks have to be carried out over longer timeframes. A defining feature of such applications revolves around the low levels of complexity encountered in such environments. Recent technological advancements in the fields of image sensors, memory technology, image data transfer, computational power, and, crucially, algorithms are making it possible to develop applications in areas marked by highly unstructured and extremely complex environments.
We see good examples of this in areas like autonomous driving and advanced robotics. Driver assistance systems became an established feature of the car industry some time ago. These systems cannot survive without a variety of sensors, some of which fall into the field of imaging sensors. These make a number of functions possible, such as lane departure warnings, traffic sign recognition, lane change assistants, blind spot monitoring, and emergency braking systems to protect pedestrians. All of these systems are extremely useful in their own right, but when it comes to autonomous driving they are absolutely indispensible. Technology has come a long way in these areas, but, at the moment, the biggest (technological) problem is still the highly complex nature of the environment and the performance of autonomous vehicles in all kinds of weather conditions, in all seasons, at all times of day, and in all kinds of traffic.
Modern robotics moved beyond the realms of pure “Teach-In” technology a long time ago. Robots are now able to monitor their working environment for possible collisions and are much more adaptive in terms of the environment and allocated tasks. In the future, robots and people should be able to work “hand in hand.” A decisive role in this will be played by sensors and, in particular, image processing.
Another example is Industry 4.0, commonly alluded to in international business by the term “Internet of Things.” This is all about the future world of big networks in which objects (things) will increasingly by equipped with sensors, actuators, and communication links. This will make it possible for objects in the real world to observe their surroundings, store and share their current status, receive instructions, and start or carry out actions. This technology will revolutionize industrial manufacturing and working environments, even in non-industrial settings, although it will also have an impact on our everyday lives. Naturally, not only will image processing play an important part in this area, it will also be considered a key technology.
It’s not difficult to identify many more areas where image processing already plays an important role and will keep moving things forward, primarily because of its economic significance. But there is perhaps one area that seems less important to some at first glance, yet it is still significant: the consumer or entertainment market. Almost all young people have instant access to image processing technology on their cell phones and no games console is the same anymore without a camera. The computing power of the components in these devices is now gigantic. They can do things like recognize gestures or offer optical character recognition (OCR), augmented reality, 3D image processing, and a lot more. In essence, it’s irrelevant whether the key driver of developments is the consumer market or industrial applications, the trend for the future in both areas is still clear. 3D applications will become more and more common. In many areas they will be standard technology and enter more new fields of application. The miniaturization trend will continue unabated, with more and more applications being made to include embedded image processing systems. Finally, technologies that are currently not common will become more important. One such technology is spectral imaging, which goes beyond just providing image information and delivers spectral information within individual pixels. Examples of applications in astronomy are now commonplace and thanks to miniaturization and significant price drops, this technology will also enter industrial image processing, where it will provide valuable additional information for use in image analysis.
There is a danger with the rapid pace of developments in image processing, the wide availability of technology in everyday life, and the way people become accustomed to new technologies: people no longer understand just how complex this field really is. To do something about this, high standards of training and continuing professional development are required in this area. Given the particularly important role played by image processing in the field of the Internet of Things/Industry 4.0, any degree with a strong foundation in engineering should include training on the fundamentals, or at least offer it to students as an option. Technology sharing at universities and research institutions also has a role to play and it can make valuable contributions to this field, which is developing extremely rapidly. There are many training, research and technology transfer establishments in Germany with outstanding experts in the field of image processing and this gives Germany a leading position in the international market. The industrial and scientific community continue to develop in this way, and it’s of supreme importance that things stay that way.
Professor Dr. Ulrich Klauck is director of the Steinbeis Transfer Center for Image Processing and Applied Information Technology, which is based at Aalen University. The Steinbeis experts at the center offer their clients services in the fields of image processing and pattern recognition, color measurement/ comparison/recognition, thermography, and high-speed imaging and image processing.
Professor Dr. Ulrich Klauck
Steinbeis Transfer Center Image Processing and Applied Information Technology (Aalen)
SU0530@stw.de