If one word is used to summarize the mobile phone ring in the past year, it must be 'Liu Haiping'. 'Suddenly like a night's spring breeze, the city is full of Liu Haiping'. From last year's iPhone X brought Liu Haiping to everyone in front of you. Liu Haiping’s controversy over the issue of U.S. and U.S. issues, and the various Android flagships are still swiftly following up, so you can see the following picture.
▲ Picture from Wei Feng Network
Throughout the history of mobile development, the most is that with the progress of manufacturing processes, the two sides of the screen are made into curved surfaces, or the screen size is elongated. However, no matter how the screen shape changes, the internal shape of the screen is almost never changed. The 'Liu Haiping' used is a major change in the internal shape of the screen, accompanied by Liu Haiping's change in UI and interaction, and even the controversy over the design of 'U.S. and Ugly'.
You know, Apple’s main new product shipments are all in the tens of millions of records in a strong product cycle, and the iPhone X is the representative of Apple’s 10th anniversary. The reason why Apple made such drastic changes this time is because Use Face ID.
In fact, the so-called Face ID should be familiar to everyone. Between a square inch at the top of the screen, an infrared lens, a pan-light sensor, and a dot matrix projector are collectively called the original deep camera system. When the user enters the face ID This system will project more than 30,000 invisible light spots on the user's face, analyze the user's face, and draw an accurate and detailed depth map for the user's face to create a unique face spectrum. Based on the original in-depth camera The collected data, the sensor will read the user's unique facial geometry. When used, it is verified by comparing the user with the previously saved data. This technique is called 3D structured light.
In general, 3D structured light can be divided into ToF (Flight Time), Speckle Structured Light, and Binocular Ranging.
Binocular ranging is similar to what we usually say 'triangulation'. Finally, we solve the triangle by imaging the spot on the camera. Finally we get the position. Due to the general accuracy, but the algorithm is troublesome, and there are certain requirements for CMOS. Therefore, it does not apply to mobile phones. The most critical thing is that there is no way to use it at night, so PASS;
▲ Microsoft's Kinect V2 application ToF
ToF (Flight Time), ToF should be the so-called 3D which is closest to everyone's understanding. In simple terms, it is to send a special signal through the infrared camera of the mobile phone. By obtaining the phase difference of the transmitted signal and receiving the signal, the depth is calculated, so that according to different Time to locate different spatial locations of objects.
However, ToF also has problems: The required camera volume is relatively large (mainly), and the algorithm is cumbersome, so the power consumption is also relatively large, and is not suitable for applications on mobile phones, the same PASS;
▲Light Code used by iPhoneX Picture from ZEALER
The last is the only light coding that is currently applied. Within a certain distance, the lattice projector emits a bunch of scattered light spots (the spot with position information), and then the user enters the first time through the calculation. The displacement of the spot gives a depth map, which is the 'standard map' entered. Then the recognition is done by comparing the initially entered face depth information map.
Compared to the other two, the accuracy and resolution of Light Coding are acceptable, and the hardware and algorithms that need to be matched are not too much trouble.
So you can see that at this stage, the cost of using 3D structured light is much higher than the traditional fingerprint recognition. So why should we insist on using 3D structured light instead of fingerprint?
In my opinion, 1, fingerprint recognition is fingerprint recognition. No matter what kind of fingerprint recognition is made, it has only one biometric function. It is only a difference in form, but there is no essential breakthrough. The 3D structure light is different. In addition to the biometric function, it can also realize the function of creating images. 2. From the interaction, pick up the mobile phone, and the mobile phone will “know” you and then unlock it. This kind of interaction is more natural and advanced than fingerprint recognition.
▲3D static modeling display
Last week, at a demonstration meeting held by the OPPO headquarters in Shenzhen, a 3D-structured optical video call display based on 5G network was demonstrated. According to OPPO, in the current environment, only 3D structured light is used as a biometric identification. Functionality, somewhat the feeling of 'big horse carts'. Therefore, OPPO combines low-latency real-time transmission of 5G networks, using 5G high-rate and low-latency features to support terminals and terminals, and real-time synthesis of terminals and cloud visual content can be achieved. 3D video, AR and VR, even apply holographic technology to mobile phones to achieve scenes that are only achieved in some movies.
Moreover, OPPO stated that it will commercialize 3D structured light technology in its mobile phone products within 6 months. If we combine the super Vooc flash chargers that OPPO has demonstrated in the past two years, the periscope cameras, the long-awaited Super Flag OPPO Find The return of the series is not just rumors. Coincidentally, Xiaomi also stated that it will be equipped with 3D structure light in the 8th Anniversary Edition of Xiaomi, becoming the first Android flagship model equipped with 3D structured light. Finally, the 3D structure light is also huge. Android ecology, the future can be expected.