The global machine vision market is still in the stage of steady high growth. The BBC Research forecasts that the global machine vision market will grow to around $ 16 billion in 2018, a compound annual growth rate of 9.2%, and $ 24.8 billion in 2023.
This growth rate may not be particularly eye-catching, but given the fact that it has been growing by an average of two digits since 2002, the strong momentum of the market is clear.
Even more remarkable is that the entire machine vision market is currently undergoing two major changes that are expected to bring about a turning point that will lead to further accelerated growth.
One of the changes is that the mainstream technology of machine vision is moving from PCs to embedded frameworks. After years of development, traditional PC-based vision systems are already quite mature and have some clear advantages. For example, high performance, the ability to implement relatively complex system functions, and the availability of richer software and hardware assets. In contrast, embedded vision systems have limited performance and are often constrained by cost, power consumption, size, and appearance. It goes without saying that there are many technical challenges to overcome. However, the ultimate goal of developing machine vision is to incorporate visual functions into various embedded systems, which makes machine vision a different life. It’s about being applicable to the side, practical and ready to use. The outlook for the “ubiquitous” market is actually quite attractive. That’s why so many players have been focusing on embedded vision in recent years and are jointly driving change in mainstream machine vision technology.
The second change in the machine vision market is driven by artificial intelligence (AI). Visual processing is already a first-class application scenario for AI technologies like machine learning. The application of machine learning and ultimately deep learning will make a big difference in the role of embedded vision products. For example, the purpose of embedded vision products will evolve from simple environmental sensing to smarter, visually guided, automated features. A typical example is the technological advancement of autonomous driving. In low-level driving assistance systems, embedded vision is typically used as a passive safety feature such as detection and reminders. In more advanced autonomous driving systems, embedded vision is deeply embedded in active safety features to turn the driver into a quick and reliable response.
These two “changes” will steadily contribute to the vibrancy of the machine vision market, with a foreseeable future and the subsequent formation of “explosion points” in other markets.
Security and Monitoring: Security and monitoring has always been a major application market for machine vision. And it has been the main driving force that has promoted the development of the industry for more than 10 years. The two new elements of growth momentum mentioned above offer great potential for this market. From the perspective of an ever-growing market, the development of embedded vision technology will increase the potential for consumer-level use such as smart door locks and video intercoms, ensuring market penetration of these applications. From a stock market perspective, security and surveillance equipment upgraded to a smarter version will have the ability to “understand” rather than just “see.” It is a step-up that can be expected to have great demand. Smart Manufacturing: Industrial applications may be considered the starting point for the machine vision market. Further use of AI technology-based embedded vision systems will inevitably make the entire manufacturing process smarter and accelerate the installation and realization of Industry 4.0. For example, robots on the production line used to be trapped in cages due to their limited visual recognition and judgment. It was taken for granted that only a constrained number of specific tasks were performed individually. A smart embedded vision will lead to the future development of collaborative robots. You will have the ability to work with humans in more complex tasks. It is the ideal form of smart manufacturing.
Self-driving: Car electronics is undoubtedly a fast-growing embedded vision market. In most autonomous driving circuits, embedded vision systems not only serve to detect the external environment in combination with LiDAR and mmWave radar, but also provide additional functionality such as fatigue testing and hand gesture recognition. It is used everywhere in the car to make the intelligent cockpit a reality.
This usage takes car safety to a whole new level and is of great interest to individual car owners. With support for advanced anti-theft systems and built-in camera surveillance, the Telematics Box provides the foundation for a smart gateway system that collects and synchronizes data from smart devices installed in the vehicle. The ability to seamlessly communicate with the vehicle’s operating system and manage and track the vehicle will be of interest to many companies.
Drones: There is a natural connection between drones and visual processing. New ways of using drones are constantly emerging. Although the market size is significantly different from that of automobiles, drones are arguably an important niche market for certain embedded vision companies.
New Retail: Amazon Go is now a pioneer in this space, using hundreds of built-in cameras and back-end vision processing systems to capture and analyze user behavior in unmanned stores. Despite the experimental nature of Amazon Go and the difficulty of deciding whether it will succeed in the future, this attempt will surely open up the retail and commerce areas of Embedded Vision.
Entertainment and Games: When it comes to using embedded vision in entertainment and games, Microsoft Kinect and Nintendo Wii come to mind first. However, these devices are just an experimental example of embedded vision, trying out consumer-level use. The ultimate goal is to find a killer application someday by perfectly combining the features of Embedded Vision with smartphones. VR / AR: After the boom we’ve experienced over the last two years, VR and AR have clearly cooled down. However, these two technologies are engraved in people’s images as future technologies. The impact of VR and AR on how people interact with machines and how information is communicated in the future is immeasurable. Embedded vision as a basic technology of VR / AR will surely not disappear soon.
The above are our conclusions regarding a survey of future embedded vision target markets. As technology solutions continue to improve and mature, Embedded Vision will surely continue to amaze us with new possibilities and unpredictable uses. Without a doubt, we are now at a turning point in exponentially increasing market growth.
