Bell Labs researchers build camera with no lens

Jun 04, 2013 by Bob Yirka report
The proposed architecture consists of two components: an aperture assembly and sensor of a single detection element. Credit: arXiv:1305.7181 [cs.CV]

(Phys.org) —A small team of researchers at Bell Labs in New Jersey has built a camera that has no lens. Instead, as they explain in their paper they've uploaded to the preprint server arXiv, the camera uses a LCD array, a photoelectric sensor and a computer to create always in-focus pictures.

Traditional cameras all use the same basic model: light coming through a lens is focused onto film, an array of or in biological models, a . This process is based on capturing the most data possible to create the best looking image. In this new effort the Bell Labs team took very nearly the opposite approach, their imaging technique is based on the idea that measurement of light as it's bounced off of an object carries a lot of redundancy—to take advantage of that, researchers use what is known as compressive sensing.

The new camera they built has just three main components: an LCD array that allows light to pass through, a RGB photoelectric sensor, and a computer to control the LCD and to process the data that is received from the sensor. To create an image, the LCD array is placed between an object to be "photographed" and the single pixel sensor. The computer sends signals to the LCD causing some of the crystals in the LCD to allow light to pass through—each serves as a tiny . The in the LCD are chosen by a —the end result is an LCD panel with a speckled pattern. The photoelectric sensor then captures the light that is allowed to pass through the LCD panel and sends the data to the computer. To create a single picture, multiple image-captures are taken with different random patterns generated on the . The data from all of the image-captures is processed afterwards and the result is a single photograph—the more image-captures taken, the higher the resolution of the final product.

Prototype device. Top: lab setup. Bottom left: the LCD screen as the aperture assembly. Bottom right: the sensor board with two sensors, indicated by the red circle. Credit: arXiv:1305.7181 [cs.CV]

The process works because the image is built from light reflected off an object as measured from a slightly different perspective. Comparing the same view as seen through many different aperture array patterns allows for building a complete picture without the need for a lens.

The upside to such a camera is its low cost—the demo made by the team at was constructed from off-the-shelf parts. Also, adding more sensors allows for creating multiple images simultaneously (three sensors allows for building three dimensional images). The downside to the process is that it takes much longer to take a picture than a lens based camera, and it only works for capturing stills.

Explore further: Computerized emotion detector

More information: Lensless Imaging by Compressive Sensing, arXiv:1305.7181 [cs.CV] arxiv.org/abs/1305.7181

Abstract
In this paper, we propose a lensless compressive imaging architecture. The architecture consists of two components, an aperture assembly and a sensor. No lens is used. The aperture assembly consists of a two dimensional array of aperture elements. The transmittance of each aperture element is independently controllable. The sensor is a single detection element. A compressive sensing matrix is implemented by adjusting the transmittance of the individual aperture elements according to the values of the sensing matrix. The proposed architecture is simple and reliable because no lens is used. The architecture can be used for capturing images of visible and other spectra such as infrared, or millimeter waves, in surveillance applications for detecting anomalies or extracting features such as speed of moving objects. Multiple sensors may be used with a single aperture assembly to capture multi-view images simultaneously. A prototype was built by using a LCD panel and a photoelectric sensor for capturing images of visible spectrum.

via Arxiv Blog

Related Stories

Sharp begins production of 5-inch full-HD LCD panels

Oct 01, 2012

Sharp Corporation has started production of 5-inch full-HD (1,080 x 1,920 pixels, 443 ppi) LCD panels for smartphones with a pixel density among the highest in the world. Production began at the end of September and full-scale ...

Recommended for you

Who drives Alibaba's Taobao traffic—buyers or sellers?

4 hours ago

As Chinese e-commerce firm Alibaba prepares for what could be the biggest IPO in history, University of Michigan professor Puneet Manchanda dug into its Taobao website data to help solve a lingering chicken-and-egg question.

Computerized emotion detector

Sep 16, 2014

Face recognition software measures various parameters in a mug shot, such as the distance between the person's eyes, the height from lip to top of their nose and various other metrics and then compares it with photos of people ...

Cutting the cloud computing carbon cost

Sep 12, 2014

Cloud computing involves displacing data storage and processing from the user's computer on to remote servers. It can provide users with more storage space and computing power that they can then access from anywhere in the ...

Teaching computers the nuances of human conversation

Sep 12, 2014

Computer scientists have successfully developed programs to recognize spoken language, as in automated phone systems that respond to voice prompts and voice-activated assistants like Apple's Siri.

User comments : 10

Adjust slider to filter visible comments by rank

Display comments: newest first

antialias_physorg
3.3 / 5 (6) Jun 04, 2013
Pretty cool. I'm almost sure one could optimize the process by not just using random settings on the LCD but settings with lowest mutual information (or lowest overlap if we just look at the 'black' pixels) for a given black/clear ratio.

The pictures in the article on arxiv look quite good for a first prototype.
210
1 / 5 (1) Jun 04, 2013
Hummmm....frame rate. Just wondering what the max frame rate COULD be for such a technology. For some reason I couldn't get the arvix article to show up...frame rate anybody?

word-
fmfbrestel
not rated yet Jun 04, 2013
No frame rate given. Must be fairly slow, but they do have a picture of a sleeping cat, so it cant be so slow that the cat would wake up. ...
betterexists
1 / 5 (3) Jun 04, 2013
"Spotlight News Stories" is missing a Sort Tab at Top Right Position......1 Tab by "Subject" Category too is needed.
Sorting by Date, Rank, Popular etc alone do NOT suffice.
Readers come from various backgrounds and they should get their own "Steak" first... then to look for anything else on the plate.
A "Subject" Tab will help in a great way....Not that Readers will stop with their preferred item.

Everyone wants Variety....But, Convenience comes first!
fmfbrestel
5 / 5 (2) Jun 04, 2013
Way to provide feedback in the place most likely to have it ignored.
Why dont you try http://phys.org/help/feedback/ instead?
beleg
1 / 5 (1) Jun 04, 2013
The LCD panel is akin to a portion of the brain.
"The crystals in the LCD are chosen by a random number generator—the end result is an LCD panel with a speckled pattern."

The crystals akin to neurons.
antialias_physorg
3 / 5 (2) Jun 05, 2013
The crystals akin to neurons.

Not really, as the state of the crystals aren't conncted (and neurons don't activate randomly for the most part)
beleg
1 / 5 (1) Jun 05, 2013
Wasn't a literal comparison. In a sense you can't process a visual event before the event occurs. Even if the 'idle' states of neurons aren't random, the source for the visual process can be regarded as random.

If there isn't any state in which the crystals aren't connected in some sense how can a pattern ever appear?
antialias_physorg
3 / 5 (2) Jun 05, 2013
Even if the 'idle' states of neurons aren't random, the source for the visual process can be regarded as random.

Not really sure I understand what you mean here. The eyes are biologically part of the brain. There's a lot of preprocessing/compression going on in the first layers after the retina before the signal goes to the cortex (edge detection, etc).
Unless you mean the object being seen as 'the source' - and that isn't random, either (this type of camera should actually have trouble seeing randomized pixel pictures,as the image is assembled by postprocessing)

If there isn't any state in which the crystals aren't connected in some sense how can a pattern ever appear?

The state of each crystal isn't dependent on the state of any other. Cameras don't make any patterns at all (they just create pixels)*.

*OK,cameras DO create patterns - but only due to inherent limitations in the entire chain from lens to sensor to memory (e.g artefacts like lens flares) .
beleg
1 / 5 (1) Jun 05, 2013
Any physical object undergoes random change. Life's building blocks are so coarse grained (inherent limitations) that noise becomes background in most instances. In a sense anything we experience is an artifact as far as time delay is concerned.

"(this type of camera should actually have trouble seeing randomized pixel pictures,as the image is assembled by postprocessing)" - AP
In randomized static pixel pictures, no. In randomized dynamic pixel pictures, yes.
You can postprocess noise. Neither will have a predictable pattern.

A random on-off switch for a cloaking device making visible objects you see invisible and reappear, I assert, will appear random to you.

The scale of noise is not limited to the molecular or smaller.
Whatever reaches your brain carries code. You can encode noise.

Lens flares are patterns, yes.