Focus: artificial intelligence, chips and other industries
Welcome all guests to pay attention to and forward
preface:
In the shadow of the decline in market sales, the pressure left on the product level has only increased. In particular, imaging functions, which are increasingly affecting consumer demand, are undergoing an unprecedented internal volume.
With the improvement of artificial intelligence, computing power and algorithms, computational images have pushed mobile phone images to new heights in recent years.
The author | Fang Wen
Image source | network
Mobile phones have completely entered the stock market
The news of cutting orders in the terminal industry chain and the weakness of the consumer layer made the Chinese mobile phone market in the first quarter full of haze.
According to data from the research institute IDC, the smartphone market shipments in the first quarter of 2022 were 75 million units, down 13.7% year-on-year, far lower than expected.
From 2020 onwards, the image capability almost occupied the highlight moment of the new product launch, and the focus of product marketing was naturally tilted towards taking pictures.
At the same time, the number of cameras is also increasing, the opening holes on the back of the mobile phone are more and more, larger and larger, quad camera, telephoto, macro, micro gimbal, periscope lens and other new technologies continue to pile up, the size of the sensor is also getting larger and larger, do our best to tap the potential of the physical space.
By 2022, the inner volume of imaging functions will be further intensified, and the innovations that were originally flagship products have been delegated to low-end products.
The progress of hardware such as lenses and sensors has slowed down significantly, and the physical space of smart phones is limited, and it is necessary to find a balance between hardware performance, space and cost, which means that the increase in hardware is bound to encounter bottlenecks.
The mobile phone image hardware stack goes inward
Image capabilities are mainly reflected in three aspects, camera modules, chips and algorithms.
The camera module includes sensors, lenses and other parts. The chip is mainly reflected in the corresponding processing power of the image chip and the CPU, GPU and NPU in the main chip.
At present, the competition for hardware is still continuing, such as periscope lenses, multispectral sensors, etc., but overall, the competition for hardware has entered a bottleneck.
Almost every time a new field is developed, manufacturers will flock to the development of related technologies, lest they be left behind.
This has also become a tacit default development direction for mobile phone manufacturers:
As long as the new technology is followed up in a timely manner, the parameters are qualified enough, and the material pile is enough, it looks like a top mobile phone in the industry, and users will buy it.
It can be perceived that once each technology approaches the ripe pole, the speed of the inner volume will be accelerated, and the image inner volume is only the embodiment of this process.
On the one hand, due to the consideration of cost and supply chain accessories accumulation; on the other hand, due to the limited physical space of mobile phones, it is necessary to balance between hardware performance and space and cost, and extreme configuration is difficult to achieve commercial success.
However, the result of such an inner volume also leads to mobile phones not being truly competitive, and once the market changes, it may be left behind.
This year, the new changes in the image battle of major mobile phone manufacturers
Algorithm tuning such as automatic HDR, super night scenes, and simulated large apertures continues to become standard, and photography has gradually changed from physical optics to a mathematical calculation problem.
Mobile phone manufacturers have more room for maneuver in algorithms and computing power, and can correct noise, brightness, color and various distortion data through high-throughput calculations, and then jump out of the inertial thinking of piling up material on the lens to solve problems.
Instead of developing more technologies or holding more patents in your hands, it is obviously a more secure way to continuously launch new machines based on mature, market-proven technologies, or fight price wars.
Starting this year, the computational photography track characterized by AI+ algorithms has become a collective breaking point for major manufacturers.
OPPO, vivo and other manufacturers have made efforts to develop their own ISP chips.
Secondly, each flagship mobile phone image has also been equipped with AI algorithms as standard to increase the permissions of the algorithm:
Huawei P50 Pro uses computational optics and primary color engines;
Honor Magic4 to The Ultimate Edition further strengthens the image engine and AI algorithm capabilities;
The OPPO Find X5 series has also begun to combine AI algorithms with ISP chips.
Some manufacturers have also gathered to find traditional camera brands to play joint cooperation, such as vivo to find Zeiss, OPPO, one plus to find Hasselblad, Xiaomi to find Leica and so on.
Vivo and OPPO's approach is similar, on the one hand, the self-developed image chip and micro-gimbal upgrade, on the other hand, the introduction of ZEISS cooperation to promote ZEISS natural color scheme, strengthen their OWN AI optimization capabilities.
From the perspective of Xiaomi 12Pro, Xiaomi's gameplay in computational photography is to introduce the seventh-generation AI engine integrated by Snapdragon 8 to strengthen its AI imaging capabilities, and cooperate with the Night Owl algorithm to create image schemes such as face recognition, all-thing focus, and intelligent frame exposure synthesis.
Apple may as early as 2022 to the iPhone equipped with a periscope camera, which can provide better optical zoom performance than the current iPhone, which will bring new breakthroughs in the hardware and software of imaging technology.
At present, from domestic mobile phones to Apple, imaging technology is taking imaging technology as the core selling point of the new flagship, and the integration of software and hardware and the development of the underlying technology based on the chip are still advancing.
New breakthroughs are expected to continue to emerge, and it is difficult to say at what stage the future of imaging technology will evolve, and it is not yet a time of conclusion.
end:
When computational photography becomes a new track for smartphone rolls, it indicates that the close-knit struggle between different manufacturers is not over, and the scramble is still a norm.
Through technological innovation to achieve high-end market breakthroughs, the smaller the window of opportunity in the back, the high-end market is so large, if today's manufacturers do not take the road of innovation, others go, and then it will be more difficult to catch up.
Some references: CBN: "Domestic mobile phone manufacturers challenge the image supply chain "deep water area"", Economic News Network: "Computing image" gradually becomes the "new track" of China's mobile phone industry", Wang Xinxi: "AI algorithm image era: Apple's opponent, changed"
The articles and pictures published in this official account come from the Internet, only for communication and use, if there is infringement, please contact a reply, we will process the information within 24 hours after receiving it.
END