Machine vision becoming the new blue sea of CIS?
In the 1980s, the research for machine vision were famous all around the world and emerged a number of application systems. After decades, machine vision has been widely used in the industrial field, which also is the famous application area for the public. With the development of intelligence and digitalization, machine vision has been utilized for tremendous application scenarios. More and more consumer products need to be upgraded with the machine vision, which has become a new blue sea for the industry.
What's the revenue of consumer machine vision market?
Around 2010, there were companies trying to apply machine vision to the consumer market, which included Microsoft, Asus, Intel, Google, and Apple. However, the deployment of machine vision in the consumer market did not go well due to the limits of technology as well as the market situation at that time.
And then, the advent of the AI era provides an opportunity for machine vision as an integral part of intelligence to enter the consumer market, vision modules are becoming more and more important for consumer electronics, and image recognition is gradually becoming a standard feature of consumer electronics in the upgrade process.
For the market, drones, sweeping robots, translation pens, intelligent boards, intelligent educational/accompaniment robots, service robots, smart dressing mirrors, smart interactive screens, SmartTV, visual doorbells, face recognition access control systems, intelligent doorlocks/doorbells, tablet PCs and other emerging consumer electronics products are need to have machine vision function.
These emerging application scenarios create a market size of hundreds of billions of yuan. According to data from Askci, the machine vision market revenue in consumer electronics and semiconductors exceeded 2 billion yuan in 2018. The rapid renewal of consumer electronics and high demand drive the machine vision market, and the market size was nearly reached 3 billion yuan in 2019.
This market trend has also triggered the industry to think about consumer machine vision. According to reports, the chip's computing power and cost, integration solutions, artificial intelligence technology, etc. are the main factors affecting the landing of machine visionin the consumer market. As consumer electronics have high requirements for the volume, price and technical maturity, consumer machine vision will not be widespread until the technologies of camera, chips and other parts of theindustry chain are fully developed.
After several years' development, the relevant chip performance has been greatly improved, artificial intelligence technologies are widely used into our lives, which also heralds the arrival of consumer machine vision applications.
CIS is the key to the consumer machine vision market
What kind of chips are needed for consumer machine vision? From the market, the mainstream consumer machine vision applications include scan code recognition, biometric recognition, obstacle avoidance applications, route planning, 3D imaging and so on. The biometric and 3D imaging markets have been particularly hot in recent years, with three main machine vision solutions: structured light, TOF, and stereo imaging.
For 3D imaging applications in consumer machine vision, ToF solutions are popular in recent years after they have been adopted by smartphones. However, it is not as hot in other consumer applications. For example, the application of the cleaning robot, ToF will be unable to effectively and accurately determine the obstacles due to excessive interference from multipath reflections in the limited space; at the same time, ToF solutions are also not applicable to open space or long distance. Also, the cost of ToF solutions discourages many consumer machine vision application enterprises. Some manufacturers choose stereo solutions in obstacle avoidance systems, and some use VSLAM to navigate and combine ToF to avoid obstacle so as to control costs.
Professionals also believed that if enterprises want to achieve higher accuracy of face recognition or high accuracy of obstacle avoidance applications, stereo and structured light may be appropriate. Compared with ToF and structured light solutions, stereo imaging has ability to resist ambient light interference and higher resolution. Therefore, stereo imaging solutions are also considered to be one of the preferred solutions for the implementation of machine vision for consumer products.
With the characteristics of CMOS image sensors, a large amount of data and information can be acquired, making it extremely effective in determining the pose of complex objects without scanning equipment assistance. And for stereo solutions that require more CIS to achieve functions such as image recognition, the performance of the CIS will affect its performance.
As a result, CIS has become one of the keys to the consumer machine vision market.
What kind of CIS is needed for consumer machine vision?
CIS has undergone a transformation from CCD to CMOS. After entering the CMOS image sensor era, back-illuminated CMOS gradually replaced the frontside illuminated CMOS and became the mainstream of the market. Then, stacked CMOS image sensors started to enter the market.
The emergence of stacked CMOS image sensors has made it easier for machine vision applied into the consumer electronics market. Stacked CIS allows image sensors to have better performance and even lighter intelligence algorithms without increasing the size of the CIS, making it more compatible with applications such as smartphones and smart imaging devices.
However, there are not many CIS that integrate global shutter function and back-illuminated pixel structure in the market, especially for the consumer machine vision market (there are some manufacturers that target back-illuminated and global shutter CIS for the industrial field.)
Faced with this market opportunity, Smartsens took the lead in developing 2nd Gen SmartGS® Technology on the basis of its back-illuminated global shutter technology. According to the website of Smartsens, the technology increases the sensitivity and dim effect by enhancing the pixel size, with an enhancement rate of up to 30%. Also, it adopts parallel voltage domain reading, which greatly improves the maximum shutter speed and perfects the clear capture of high-speed moving objects. Based on a new capacitive process, 2nd Gen SmartGS® Technology can reduce read noise by more than double, while both area HDR technology can greatly improve the dynamic range in specific locations, solving problems such as overexposure of signal light imaging for night photography.
Backed by multiple generations of SmartGS® Technology, the SmartGS® series products have become the newest addition to the consumer machine vision market with their outstanding imaging performance and hardcore technical strength. It is understood that the sales of SmartGS® series products have tripled between 2019 and 2020. This year, the SmartGS® product line is the highlight of the year - the world's top one in consumer machine vision shipments.
As a part of the domestic chip market, the achievements made by Smartsens in the consumer machine vision industry provide a boost to the development of the local semiconductor industry chain. Furthermore, the achievements of Smartsens in this emerging field indicate that the influence of local semiconductor companies is expanding.
As Yole pointed out in its report, the next era will involve robotics, augmented reality (AR) and the intelligent Internet of Things (IoT). China is well positioned in these areas, which provides opportunities for local CIS companies to grow. As a result, CIS rankings may also change with these new changes.
The consumer machine vision market is an emerging field. Consumer electronics have long existed in all corners of our lives, they are made different by the empowerment of machine vision. The market of machine vision will take off in the complex consumer application scenarios. Let's expect the next hot topic and application in the industry.
Follow us on social media to get insights of China's optoelectronic market.
Scan to follow CIOE English WeChat