Twitter
Advertisement

This pocket camera can take on DSLRs

An Indian researcher in the US develops a ‘multi-eyed’ camera that uses smart software to capture pro quality photos.

Latest News
article-main
This tiny camera uses multiple lenses to capture pro-quality shots
FacebookTwitterWhatsappLinkedin

With many phone manufacturers these days, the name of the game is ‘a camera that rivals DSLRs’. From the big brands to the legions of Chinese manufacturers, each is vying for photography chops that prompt you to ditch your regular camera altogether. But if this tiny company from Palo Alto, California, has its way, they just might be the first to make it happen.

Known simply as ‘Light’, the company has invented an admittedly alien-looking camera clearly packed to the gills with multiple lenses, which teams up with advanced algorithms that deliver pro photo level shots. Best of all, it accomplishes this feat using good old physics along with smart image processing, instead of resorting to quick fixes and workarounds that many current-day smartphone cameras tend to employ.

The brains behind this camera is Rajiv Laroia, an IIT Delhi alumnus who has in the past played R&D roles in fields including LTE networking and optics development. The camera that he and his team have developed is poised at bridging the challenge of high photographic quality with pocketable portability.

Where conventional DSLR cameras have large image sensors and lenses containing plenty of high-quality glass, these very factors make them both bulky and expensive. Cell phone cameras on the other hand use plastic lenses that team up with much smaller sensors, meaning they are able to receive only a small amounts of light from a scene that result in grainy pictures--especially in low-light environments. Also being housed in space constrained smartphones, these cameras have fixed focal lengths and tiny apertures that stifle artistic expression. So pushing photography limits with a camera phone implies using techniques like image recognition to create blur effects, in a bid to simulate the shallow depth of field generally produced by conventional DSLR cameras.

Not so with this new device--as evident from the photo, the camera consists of what appears to be a spider’s head of eyes, 16 in all, where each lens has its own 13MP sensor. The camera effectively captures a scene from all these sensor simultaneously, where these multiple photos are combined digitally from groups of lenses to create a single composite frame. Apart from letting the photographer select the zoom in the final photo, she can also select aspects like depth of field even after the shot is taken.

These 16 individual lenses are grouped across three different focal lengths—five are 28mm equivalent, five are 70mm equivalent, and six are 150-mm equivalent and all of these lenses have an aperture of f/2.4. The magic, however, lies in the way these lenses are arranged.

Five of these lenses point straight out of the camera, while the remaining 11 can actually point sideways. Attached to each lens is an element that moves it back and forth for focusing, but with the sideways lenses there is an additional mirror that can be moved to capture a straight or a side shot.

When capturing a photo the camera’s algorithms select 10 of the 16 camera modules depending on the zoom level selected, where the component images are combined into a single large image. For example when capturing a photo at 28mm (such as a wide angle shot of a landscape,) several lenses jump into action: all five 28mm cameras capture the scene albeit with each doing so from their slightly different perspective, all five 70mm lenses also capture a shot where each captures a quarter of the scene due the focal length’s zoom. Finally a fifth 70mm module pointing at the centre of the frame captures extra detail from the center of the image.

Image data from all of these cameras are combined using an algorithm that extracts scene details based on the optics and viewpoint of each lens. The resulting combined photo is up to 52MP in size. Interestingly this image combination can be done either on the phone or on a computer running their software, which uses the same algorithm.

A similar process happens when capturing a 70mm shot or a 150mm zoom shot. For in-between focal ranges the camera captures a wider field of view using the technique above, then crops the photo to match the required focal length.

This camera will use off-the-shelf parts including plastic lenses currently deployed in smartphones, conventional image sensors, even a Snapdragon 820 processor that does all of the processing within an Android operating system. The Light L16 camera described here is expected to be commercially available for $1,699 (approximately Rs 1,13,485) by early next year, and the company aims to propagate this technology into affordable smartphones in future.

Find your daily dose of news & explainers in your WhatsApp. Stay updated, Stay informed-  Follow DNA on WhatsApp.
Advertisement

Live tv

Advertisement
Advertisement