Xiaomi officially published an article on the Mi 12 series of imaging systems in detail. The Mi 12 series focuses on solving the problem of not taking pictures fast enough in the past. At the same time, Mi 12 Pro first released the Sony IMX707 sensor, which has better night scene shooting capabilities.
JOIN US ON TELEGRAM
Moreover, Xiaomi said that it took two years to rewrite the entire camera architecture and integrate software and hardware capabilities to create a new Xiaomi imaging brain. The Xiaomi imaging brain on the Xiaomi Mi 12 series includes a total of intent recognition units, computing units, and acceleration engines. The four parts of the ecological engine.
-
The purpose of the intention recognition unit is to determine the user’s intention and then dynamically adjust the intensity of the algorithm, including intelligent dynamic allocation of memory and so on. When the user keeps pressing the shutter, Xiaomi Imaging Brain will think that the user’s intention is to record more images, and then it will give priority to ensuring the speed of taking pictures.
-
The acceleration engine uses the Snapdragon 8 platform for empowerment, integrating, and fully scheduling the computing capabilities of multiple modules such as CPU, GPU, DSP, and ISP, allowing the entire process to be coordinated in parallel, so as to achieve the optimization effect of greatly increasing the speed.
-
The computing unit uses powerful computational photography capabilities to call advanced algorithms to optimize image quality, such as multi-frame synthesis, noise reduction algorithms, anti-false color algorithms, white balance calibration, etc., to make the picture clearer and more natural. So as to achieve the effect close to or even beyond the traditional optical photography.
-
The ecological engine opens up the pipeline from hardware to third-party applications and opens up the capabilities of Xiaomi’s image brain to third parties through Xiaomi’s own SDK so that third-party applications can also use the capabilities of Xiaomi’s image brain in the future to achieve image quality adjustment. Excellent, the effect of speed improvement.
Officials said that the Mi 12 series cameras startup 21.2% faster, take pictures 53.7% faster, continuous taking pictures are 209% faster, and night scene taking pictures is also 39.9% faster. Among them, the heterogeneous computing and algorithm parallel processing of Xiaomi’s image brain plays a key role.
In order to optimize the shooting process, Xiaomi adopted a new ” thread parallelization technology ” to move the processing algorithm upwards while processing multiple photos in parallel.
In addition, the original hardware HAL layer is abstracted and managed as Xiaomi’s own HAL layer, allowing hardware algorithms to be combined and changed at will, and even plug-in management.
Coupled with algorithm preloading optimization and algorithm concurrent optimization, the ability to handle multiple continuous shooting is greatly improved. Heterogeneous computing can reduce unnecessary algorithm library loading, thereby achieving the effect of reducing memory and speeding up speed.
Furthermore, the Sony IMX707 CMOS sensor, which was first released by Xiaomi Mi 12 Pro, is 1/1.28 inches in size, has 50 million pixels, and has a single pixel size of 1.22μm. It supports four-in-one pixels. Compared with the IMX766, the sensitivity of the sensor is increased by 49%, and it can achieve a pure night scene shooting effect with Xiaomi’s self-developed Night Owl 2.0 algorithm.
Not only that, but the enhancement of light sensitivity also enables the IMX707 sensor to achieve faster camera speeds and better capture capabilities.
(via)
