This week, Xiaomi Mi 9 set off a large-scale warm-up activity. At the same time, some executives are also “truss” with glory, including the TOF deep-sensing lens that slams the latter’s V20 as “scratch” and “swindle users to spend money.”
In the early morning of this morning, Xiaomi Product Director Thomas wrote a long article on the principles and current application of TOF technology.
Wang Teng mentioned that TOF has application imagination space in AR games, 3D modeling and face recognition, and OPPO R17 Pro is the earliest application-listed mobile phone, but it has not been widely publicized. He revealed that the Xiaomi Mi MIX 3 pre-research machine based on TOF technology completed the function verification in Q3 last year, but decided to continue the pre-research based on the low resolution of the sensor and limited application scenarios, and did not introduce mass production. Presented when giving users value.
The following is the full text:
Talk about TOF, taking into account not to waste everyone’s time, the length of the text as much as possible to control the short and popular, there is a need to elaborate on the comments we put in our comments or post.
From the principle, device composition, advantages and disadvantages analysis and application scenarios, the following dimensions:
1, TOF principle: In fact, it is not complicated. The laser emitter emits a light pulse. After encountering an object, the light will be reflected. The captured light passes through the captured time, which is the time of flight. The object and lens can be judged by simple formula calculation. Distance between
2: TOF component: The main component consists of three parts: the light source (here the laser emitter), the lens and the photosensitive element.
3, several advantages and disadvantages of the ToF program (mainly compared to the other two mainstream 3D solutions, see figure)
Advantage 1: First, the working distance is long, and the effective & real-time depth information within 5m can be obtained;
Advantage 2: Wide application range, no matter whether the object has feature points or not, whether the ambient light is strong (such as daylight) or weak, the effective depth of field information can be obtained;
Advantage 3: The accuracy of the long distance is high. The absolute precision of the ToF in the mobile phone and the object, that is, the relative precision between the objects, can reach the level of a centimeter.
Disadvantage 1: The current mainstream ToF sensor available on the mobile phone has a relatively low resolution (180*240, 240*320, 240*480, etc.), so the accuracy and X/Y resolution at close range will be relatively low. Interested in knowing the accuracy of the pre-structured light;
Disadvantage 2: The power consumption and heat generation of components are relatively large at work. Long-term work requires good heat dissipation conditions, and it needs to be continuously optimized for use on consumer electronic devices.
Disadvantage 3: At present, the solution based on ToF solution is not fully mature, and the corresponding content production and development groups are weak, and the supported application scenarios are few.
4, the main application scenarios of TOF
TOF has some applications in the industrial field, for example:
In the logistics industry, the hot robot, TOF is applied to the robot to help identify the object, can be used for auxiliary packing, box packing, box stacking, box marking and so on.
At present, the OPPO R17pro is the first to be listed on the mobile phone. It is limited by the accuracy and power consumption mentioned above. There are not many applications. Of course, OPPO is also trying to do it without the main promotion, and even the official website has not introduced it.
Now I see the mainstream in the industry in two directions: applications based on depth perception, such as AR special effects games, ranging, 3D modeling, etc., in addition to biometric recognition based on 3D information, such as TOF face recognition instead Front structure light.
The imaginable space is large but it takes a step by step to solve various problems and develop corresponding applications.
5, finally said the progress of Xiaomi, we recognize that this is a big industry direction – based on 3D information in the future can do a lot of exploration, and invest in research and development resources in the last Q3 based on MIX 3 pre-study prototype and transfer completed functional verification (see Figure), but based on the lower resolution, limited application scenarios and other reasons, decided to continue to do pre-research, temporarily not introduced mass production. We hope to present it to you at a suitable time in the future with more complete user experience and truly bring user value.