ToF high-performance development platform, what are the implementation scenarios of ADI innovative applications?

Many people play games, so do you know about ADI? Gamers perform various actions in the air with their bare hands in front of game equipment, easily playing game roles in various virtual worlds; various robots navigate autonomously in various scenes, easily avoiding It can quickly move around obstacles to achieve various specific target tasks such as autonomous cleaning, express delivery or inspection; when truck drivers are drowsy in the afternoon, they stop the vehicle to rest for safety under constant warning sounds from the cockpit... These are all There are more and more real-life technology application scenarios in our lives, and ToF (Time of Flight) is becoming one of the key enabling technologies for these many innovative applications.

ToF has quickly entered the public eye from its initial trendy application on mobile phone cameras, and has begun to show its prominence in various application fields. According to the IHS Markit report, based on the various advantages of the ToF solution, the ToF market size is expected to reach US$1.5 billion in 2022, accounting for about 50% of the 3D sensing market. ADI Industrial Marketing Manager Li Jia also believed in a recent interview with the media that as one of the three mainstream solutions in the field of 3D depth vision, ToF technology is not only used on mobile phones, but also in VR/AR gesture interaction, automotive electronics ADAS, Many fields such as security monitoring and new retail have begun to show their talents, which is expected to promote a new wave of 3D intelligent sensing application innovation.

The ecosystem is booming, promoting a new wave of 3D intelligent sensing applications of ToF technology

As we all know, the biggest advantage of ToF in comparison with binocular and structured light is distance measurement. Its depth calculation accuracy often does not change with distance and can basically be maintained at the centimeter level. Therefore, ToF is very suitable for scenes containing large-scale movements. Before 2010, ToF technology was mostly used in scientific research fields such as cosmic measurement and high-precision microscopes. With the substantial improvement in the performance of light-emitting components, industries such as autonomous driving, VR/AR, industrial robots, and the Internet of Things have gradually emerged, further driving the ToF Application market development. “For now, the application scenarios with the highest ToF matching include ranging and perception of the driving environment in autonomous driving, monitoring of safe distances for human-machine collaboration in the industrial field, machine vision, volume and calculation in the logistics industry, robot navigation, etc. , all taking full advantage of the ToF sensor's advantages in precise positioning and distance calculation." Li Jia pointed out.

Pulsed ToF is one of the depth-of-field measurement techniques. In order to accurately synchronize light pulses, these calculations need to be performed on millions of pixels per second and adjusted according to operating conditions, which is extremely difficult for mixed-signal circuit design and applications. challenging, ADI is one of the few companies with the technical expertise to provide a ToF solution that is both high-performance and cost-effective. In 2014, ADI began to customize ToF technology for a well-known foreign AR glasses. The product was successfully implemented and realized human-computer interaction and 3D reconstruction functions; in 2016-2017, ADI ToF technology was used for in-car gesture recognition; in 2018, ADI ToF The technology began to cooperate with a certain domestic brand of mobile phones and was successfully launched in batches... "So far, ADI's ToF technology and products not only cover industry, consumption, automobiles, etc., but the boundary expansion of applications is still continuing. There will be more space and more in the future." The combination of high-resolution depth data with powerful classification algorithms and AI will unlock more new scenes," Li Jia said.

In the past period of time, ToF applications have gradually shown a trend of accelerating their implementation. For example, SLAM (Simultaneous Positioning and Mapping) has begun to play an active role in the autonomous positioning and navigation of robots, effectively promoting the realization of human-computer interaction; the ToF module developed by ADI combines image sensors and VGA ToF sensor modules with built-in images The processor solution provides a wider range of collision detection and prevention for applications such as car reversing systems, door opening protection systems, parking assistance systems, and blind spot detection. "To further enhance the large-scale application of ToF technology in the industry, we always need the cooperation of partners in the upstream, midstream and downstream industry chain, including optics, electricity, chip level, circuit level and module factories. There are even some applications that require professional teams. Develop and adapt the entire solution." Li Jia pointed out. It is reported that ADI has launched a 3D time-of-flight development kit with Arrow, and has cooperated with a number of software solution companies such as Vision Intelligence to launch complete solutions that meet a variety of application scenarios.

ADI ToF analog front-end core chip analysis to build an accurate measurement hardware platform

Generally speaking, ToF technology can be divided into two types according to different modulation methods: pulse modulation and continuous wave modulation. Both ToF modulation methods have their own advantages and disadvantages. Points that need to be comprehensively considered based on actual application use cases include measurement distance, system environment, accuracy requirements, thermal/power consumption limitations, form factor, and power supply issues. According to Li Jia, ADI's ToF solution currently mainly uses the pulse principle of CCD sensors. As a global exposure device, CCD has very good performance in outdoor long-distance scenarios. At the same time, CCD, as a sensor for global exposure and narrow shutter exposure, has very strong anti-interference performance against external interference.

Specifically, ADI's ToF solution uses a high-performance ToF CCD and integrates a 12-bit ADC, a depth processor (processes the original analog image signal from the CCD into depth/pixel data), and a high-precision clock generator (for ToF analog processing front-end for CCD and laser generation drive timing). The ToF CCD main chip ADDI903x series has very good performance. The sensor converts the collected optical signals into electrical signals, uses depth calculation to send the final distance information to the SOC for processing, and then provides the data to algorithm adaptation. The whole process involves sensors and lenses. , transmitting devices, platforms, applications, etc. For hardware manufacturers, they need to cooperate with module manufacturers to create small, highly integrated products to adapt to application manufacturers and then adapt to different application scenarios.

Because the ADI ToF system-level solution uses a ToF image sensor with a resolution of 640×480, its resolution is 4 times higher than that of most other ToF solutions on the market. At the same time, this system-level ToF solution can provide higher system performance than CMOS solutions at the same size or the same cost. For example, high resolution can better distinguish the subject from the background in an environment with complex lighting. Thanks to the CCD architecture designed for the 940nm luminescence band, ADI ToF solutions can also more accurately capture images in sports environments.

In addition to front-end chips and sensors, more optoelectronic devices are also required according to the needs of CCD, and the peripheral devices of ADI's ToF solution are all products of ADI. "ADI's ToF system-level solution does not recommend that customers make too many modifications when using it, because each module in the system is the best overall performance state of the system that ADI has achieved through long-term running-in with its partners." Li Jia emphasized in particular road.

Find and solve potential technical pain points and explore wider paths for ToF applications

More and more application scenarios have opened up one outlet for ToF, but related challenges may also arise, such as how to avoid interference between multiple ToF application terminals in the same scenario, especially for majors such as automobiles and industry. level scenarios, the design of the ToF sensing system not only needs to strike a balance between accuracy, range, response time, resolution, cost, power consumption and available packaging requirements, but also needs to address various uncontrollable problems that occur in different actual situations. Factors, some customized designs should be made for the flexibility and anti-interference of the sensing system, such as adding some high-reliability filtering and anti-interference devices and modules, and loading relevant software algorithms to ensure that the system has sufficient Ability to deal with different types of emergencies.

Imagine if multiple autonomous robots are sorting goods in the same large warehouse, or two self-driving cars approach an intersection at the same time, and ToF cameras cannot eliminate interference from mutual light sources, then the application range of using ToF technology for accurate depth measurement will be Severely limited (ADI’s development roadmap is expected to achieve simultaneous short-range detection scenarios of 64 cameras). Even on the consumer side, as various applications increase, interference between devices will become a real problem. Therefore, ADI believes that the ability to prevent interference or eliminate interference in ToF systems will become increasingly important. According to reports, ADI currently uses a patent-pending algorithm that can avoid or eliminate all irrelevant light information and only use the light information of its own laser source, so it can give correct depth information. The depth processor uses a pseudo-randomization algorithm and special image processing functions to eliminate multi-machine interference. Therefore, multiple ADI ToF systems can be used in the same environment.

In addition, as 3D algorithms further mature, data analysis will be used to collect a large amount of effective information about people's behavior. This technology may first be used in building control applications, such as access control systems. Vertically mounted sensors add depth information, meaning people can be counted very accurately. Another use case is smart automatic door opening, which can distinguish between people and only open when a real person is detected. ADI is also currently developing software algorithms for people counting and human differentiation. By using depth information, people can be classified with high accuracy in many challenging conditions, such as in dimly lit environments or with no ambient light, in populations. In densely populated areas, and in situations where personnel are wearing complex clothing. Best of all, people counting errors are virtually eliminated.

Summary of this article

For many technologies, the maturity of the technology is not the only factor in the popularity of the technology, but also has a great relationship with the cost of the technology and the coordination of the industry chain. In the past 2019, ToF technology has been widely adopted by leading mobile phone manufacturers, taking a key step in the evolution of technology.

How to use technical means in the future to truly achieve a balance between cost, power consumption, volume, speed, lifespan, stability and anti-interference ability, etc., to achieve a more optimized level than the current level, and then realize the practical application of ToF technology The doubling of reliability is a prerequisite for the popularization of ToF technology and even the healthy development of the entire market. It is also a research focus that ToF technology system-level solution providers such as ADI need to consider. The above is the ADI ToF high-performance development platform and the relevant technical analysis of ADI innovative application landing scenarios. I hope it can help you.