How phone cameras will become better
Gone are the days when you had to spend upwards of Rs50,000 to get a good camera smartphone. Smartphone cameras have improved a lot in the last few years. Even the budget smartphones are no longer saddled with poor quality sensors. Phone makers are working on fine tuning the algorithms that power the camera app on the phones. The likes of Huawei roped in well-known photography brands such as Leica to improve the photography performance on the Huawei P9 phone. Dual camera is another recent advancement in camera technology which is being used to enhance depth in images. Some of the recent announcements suggest that this is just the beginning. We take a look at some of the upcoming camera technologies which you will get to see in some of the upcoming smartphones.
Google has developed an AI (artificial intelligence)-based algorithm which can improve images even before they are clicked by a smartphone camera. It was developed in collaboration with researchers from Massachusetts Institute of Technology (MIT) and will be showcased at the SIGGRAPH digital conference in Los Angeles this week.
The algorithm works by sending a low-resolution version of an image (which the user intends to capture) to a web server which responds by sending back a three-dimensional grid. Every cell in the grid contains a set of formula which determines the changes in the colour values of the image. When the grid is applied to an image, the formulae will decide which part of the image needs retouching or colour enhancement.
The AI has been trained in a dataset of 5,000 images created by MIT and Adobe Systems. What sets it apart from other photo-based algorithms is the small size. It takes up as much space as a single digital photo, and will run on smartphones without lags, according to MIT researchers.
Android smartphones running too many processes often tend to slow down. Camera is a resource-intensive tool and if your phone’s memory is occupied elsewhere, it is likely to take longer to process images. Sony has found a way around it by developing a 3-layer CMOS sensor for smartphones which comes with a dedicated DRAM (dynamic random access memory). It is placed between the BSI (Backside Illuminated) sensor and the circuit layer responsible for signal processing.
DRAM is a type of memory which stores frequently-accessed programs or data separately so they can be quickly accessed by the processor. Adding a DRAM will improve the speed of image sensors and will result in faster image processing. The technology was unveiled at Mobile World Congress (MWC) 2017 earlier this year. There is no update yet on which upcoming Xperia smartphones will have this feature.
Smartphones with higher optical zoom are too bulky or have a lens which protrudes like a point-and-shoot camera. Oppo has come up with a solution which uses the phone’s width rather than depth to incorporate a 5X optical zoom lens. Showcased at MWC 2017 earlier this year, the technology is based on the design of a periscope. The angle of the telephoto lens has been shifted by 90 degrees and a prism has been added to change the direction of the light entering the camera by 90 degrees. So instead of cropping the image, the camera sends it through the telephoto lens when the user zooms in on the image. This will minimize loss of clarity in landscape shots. Asus did something similar with the ZenFone Zoom, but was able to muster only 3X optical zoom.
Samsung is working on improving the design of the camera sensors. Its new generation of camera sensors, announced at the recently concluded Shanghai edition of MWC, are coined ISOCELL as they are based on Samsung’s proprietary ISOCELL technology. The sensors work by separating each pixel with a physical wall which minimizes colour crosstalk among pixels. This results in images with more colour accuracy even if the pictures are taken with low-resolution cameras. It is rumoured to show up in the Galaxy Note 8.