The newly released Huawei Mate 70 Pro and Pro+ also feature four cameras, but the additional one is special: Huawei calls it the “Red Maple Original Color Camera,” which is essentially a miniature multispectral spectrometer.
Multispectral imaging devices are mainly used in professional settings like agriculture, food inspection, and other specialized fields. It’s rare to see them in smartphones, and a lens with such high specifications as Huawei’s is almost unprecedented.
But in the near future, similar multispectral sensor modules will appear in more smart devices around us.
Solving the “Color Blindness” Problem in Photography
Although smartphone manufacturers have been competing on imaging capabilities for many years, with phone camera modules getting larger and more prominent, there’s still one tough nut to crack – color reproduction.
Currently, traditional phone camera sensors typically capture colors through red, green, and blue (RGB) channels to present colors within a certain range. Each pixel only contains color information for red, green, or blue, requiring algorithms to estimate the true color of each pixel point by analyzing surrounding pixels’ color relationships, ultimately attempting to reproduce what we see with our naked eyes.
▲ Each pixel is responsible for recording information of a single color
However, colors composed of red, green, and blue only represent 75% of human-visible colors, resulting in significant color loss in photos.
Since it relies on algorithmic “estimation,” accuracy can’t be fully guaranteed. In more complex color and lighting environments, algorithmic accuracy decreases further, explaining why skin tones sometimes appear yellowish or reddish in photos.
To address color inaccuracies, manufacturers have attempted solutions from both software and hardware perspectives.
Google Pixel, known for computational photography, introduced the “Real Tone” color algorithm to restore authentic skin tones for different ethnicities during photo post-processing.
Many manufacturers have also tried switching tracks, not pursuing perfect color reproduction but adopting a stylized approach to achieve better visual appeal rather than 100% accuracy.
Huawei has been incorporating color temperature sensors since the P20, starting with a 5-channel sensor and upgrading to the P50 series’ 10-channel multispectral color temperature sensor, capable of capturing wavelengths beyond standard RGB, including some non-visible spectrum bands.
Then, through the XD Fusion Pro algorithm, it merges visible and non-visible light to present more color details.
Using multi-channel multispectral sensors to correct camera “color blindness” is actually an industry mainstream solution, with vivo and OPPO also experimenting with similar multispectral color temperature modules in their flagships.
Simply put, these sensors can capture more specific wavelength bands of color but can only sense and judge overall environmental light, without providing localized light source and color information or imaging capabilities.
The 1.5-megapixel Red Maple Original Color camera module essentially contains 150 independent spectrometers, offering higher color sensitivity and imaging capabilities, simultaneously capturing 2D images and spectral information, with recording capabilities far exceeding RGB three-color.
▲ Spectricity also produces similar multispectral cameras
In plain language, while the Huawei Mate 70 Pro’s main, telephoto, and ultra-wide cameras take normal high-definition photos, the Original Color camera specifically records color details of each area in the photo, then the processor retouches the photo according to this precise color information.
But the potential of this high-spec multispectral camera goes far beyond this.
The Phone’s “Fire Eyes”
Different materials absorb and reflect light differently, presenting different spectra.
Simply put, different materials show different “colors,” some visible to the naked eye, others requiring multispectral cameras to capture.
Through these spectra, we can deduce the properties and states of materials, which is why multispectral devices frequently appear in environmental monitoring, agriculture, medical diagnostics, chemical analysis, forensic investigation, and other professional fields.
Industry insiders told iFanr that spectral modules in the industry can achieve up to 2 megapixels, while the multispectral camera on the Huawei Mate 70 series reaches 1.5 megapixels.
While this specification might not be sufficient for the professional fields mentioned above, if Huawei opens third-party interfaces, it could enable the consumer-focused Mate 70 to create more possibilities.
For example, we could use the phone’s spectrometer to scan agricultural products from markets, detecting pesticide residues invisible to the naked eye through different spectral characteristics. Similarly, it could be used to detect fluorescent agents in tissues.
▲ Using hyperspectral cameras to identify moldy plums and foreign objects, source: advian
Skin conditions can also be photographed with the spectrometer, identifying issues like dryness, brightness variations, dullness, and spots early for better skin management.
Since blood exhibits near-infrared and mid-infrared spectra that can theoretically be detected by the spectrometer, it might be possible to take a “selfie” with the Huawei Mate 70 series for rough monitoring of body metrics like blood sugar and oxygen levels.
In essence, this multispectral camera truly functions like Sun Wukong’s “Fire Eyes,” capable of revealing objects’ true nature simply by “looking.”
For AI, this is absolutely a dream capability.
Currently, AI assistants mainly interact with the external world by analyzing images captured through cameras, but the information from a flat image is limited, making it difficult for AI to perform extended interpretations.
▲ iPhone 16 “Vision Intelligence”
However, if multimodal AI could access multispectral cameras, it would greatly enrich the external information it can obtain. For instance, AI could “look” at a refrigerator and analyze ingredient freshness based on spectral characteristics, providing more appropriate and detailed dinner recipe suggestions.
Beyond phones, spectral cameras in smart home devices also have broad potential. Imagine an oven that can intelligently judge ingredient conditions and decide the optimal cooking method without manual temperature and time settings.
Using phones to photograph dogs for smart recognition might be criticized, but if it could scan dogs for quick skin disease checks, that would truly change lives.
More Devices Are Coming
Why has such a useful multispectral device only now made it to our phones?
Traditional spectral instruments’ size and weight were far from being compressible to phone camera module dimensions. The core diffraction component – the grating – in spectral devices is difficult to miniaturize and requires certain transmission ranges. Sometimes multiple gratings are needed within the device for broader spectral ranges.
▲ A spectrometer principle diagram, with the grating in the middle
Domestic optical enterprise Seekoptics explored using algorithms to replace physical diffraction components, successfully miniaturizing spectrometers.
▲ Seekoptics’ multispectral camera
Another factor hindering multispectral cameras’ integration into smartphones was computing power: when pressing the phone’s shutter, the processor needs to process photos using detailed spectral information. Insufficient performance would significantly lengthen photo processing time, seriously affecting the shooting experience.
This problem was resolved as phone processors integrating CPU, GPU, and NPU (AI processor) continued breakthrough performance improvements.
In 2022, Huawei Hubble invested in Seekoptics, and two years later, the first smartphone with a 1.5-megapixel multispectral device – the Huawei Mate 70 series – was born.
But Huawei won’t maintain this lead for long. In fact, overseas companies are exploring similar paths to introduce multispectral chips into mobile devices.
Belgian company Spectricity launched the S1 multispectral sensor specifically designed for smartphones in 2023, with principles and effects similar to the Red Maple Original Color camera, capable of improving color accuracy in phone camera outputs.
▲ Spectricity’s multispectral camera module
This year, Spectricity announced a partnership with Qualcomm, aiming to develop excellent spectral image sensor products for the Snapdragon platform.
While no consumer phones featuring the S1 have been released yet, Spectricity states they’ve partnered with several phone manufacturers, including Chinese brands, with related products launching soon.
Perhaps next year, this sensor, currently unfamiliar to the public, will frequently appear at various flagship phone launches, becoming another novel selling point.
For consumers, this camera not only means more accurate color photography but also contains new possibilities for revolutionizing human-computer interaction.
Reference Articles: “Spectral Camera, Huawei Mate’s Most Outstanding ‘Smart Eye’ in History” “Spectral Imaging, Entering a New Era of Phone AI Imaging”