Qualcomm will significantly improve Android phone cameras in 2018
- Supreme Court rejects Essar’s plea against Hazira port expansion
- HDFC Property Fund aims to raise $500 million from overseas investors
- Essar Oil seeks loans from traders as banks fear Russian links
- CoLive raises Rs12 crore from SAR Group and others
- Cybercrimes cost firms $600 billion last year: McAfee report
It is unlikely that many Android phones which will be on sale next year will still have a single camera setup. Chip-maker Qualcomm has unveiled the Spectra-embedded image signal processors (ISP), which focus extensively on depth sensing. A company better known for making the chips that power a lot of the Android smartphones already, its new ISPs are a part of the Spectra Module Program.
There are three modules in all—one is an iris authentication camera, the second is a passive depth sensing module and the third is an active depth sensing module.
The iris authentication module will embed the biometric security feature in smartphones. This isn’t exactly new (though it is still relatively uncommon), iris authentication is already available in the Samsung Galaxy Tab Iris launched last year and certified for Aadhaar use, and also in the Galaxy S8 and Galaxy S8+ smartphones this year. Google also has the Project Tango, though that is hindered by the need for a large camera sensor, and Intel has the RealSense technology that did make its way into the Dell Venue 8 7000 tablet but hasn’t really caught on since. However, Qualcomm’s evolution comes in the form of better detection and more security. The new iris module has a front sensor module with 1.12um pixel size, and will be able to identify you even if you are wearing sunglasses and cannot be fooled by a photo or a 3D model of someone’s face. The latter is because of the new liveness detection algorithms.
The passive depth sensing module will bring basic depth sensing capabilities to complement the camera, in most entry-spec and mid-range Android phones. The module, paired with the camera, will be able to create two images of the scene you are capturing, by sensing the distance between two objects.
Then there is the active depth sensing module, which will include an infrared illuminator as well as an infrared camera. The illuminator will project a laser light pattern, while the camera will read the pattern to understand any additional elements in that path.
With the Spectra 2nd generation image processor, Qualcomm claims to have improved noise reduction, made electronic image stabilisation faster, motion compensated temporal filtering in video capture for smoother videos particularly when it comes to handling lighting and face detection as well as bokeh photography will rely on machine learning-assisted computer vision.
The enhanced depth sensing capabilities open up the doors for very useful applications apart from just improved photography and video recording capabilities. The real-time depth mapping, tracking hand movement and object detection will be very relevant for virtual reality (VR) and augmented reality (AR) applications. Qualcomm has already made a big push into VR with the VR 835 platform, which is designed for standalone VR products that don’t require a PC or tethering to a phone. While they may not be as powerful as perhaps the Oculus Rift or even the HTC Vive, they could theoretically take advantage of the Spectra 2nd generation too for an experience that could perhaps be better than what Google’s Daydream or Samsung’s Gear VR currently offer.
The timing of Qualcomm’s announcement is quite interesting though. Face recognition, iris scanning and augmented reality are expected to be big with the upcoming Apple iPhone expected in September this year. If this has to pit Qualcomm against Apple next year, the chipmaker thought it would perhaps be a good idea to fire the first shot.