DARPA tests breakthrough camera tech

The scientists at Defense Advanced Research Projects Agency want to build high powered scopes and cameras for soldiers and Marines up to 2,000 times more powerful than current imaging capabilities.

A recent test found a camera could accurately read signs and license plates up to 270 yards away. Run under the Advanced Wide Field of View Architectures for Image Reconstruction and Exploitation (AWARE) program with Duke University, the recent test used a 1-gigapixel camera made up of 100 micro-cameras.

DARPA leaders want to eventually build a 50-gigapixel camera that would usher in the imaging capabilities that are 2,000 times more powerful than today’s cameras. The next step is to build a 10-gigapixel  camera capable of reading license plates 540 yards away.

Scientists and engineers understand the camera is no good to soldiers on the ground if it can’t work at night or in bad weather. These super cameras are being built for both.

DARPA officials have said they have seen major breakthroughs with their High Operating Temperature Mid-Wave Infrared (HOT MWIR), which will allow engineers to build more powerful  hand held thermal imagers and long-range thermal scopes. The HOT MWIR developments also fall under AWARE.

Research has found the “advances in cooling, novel high operating temperature detector design and small pixel spacing allow for a large format sensor in a small, low power package,” according to a DARPA statement.

“Never before has a MCT MWIR with ‘see spot’ capability been developed into such small handheld sights and potentially unequaled performance in future sniper scopes,” said Nibir Dhar, the DARPA program manager for AWARE. “The HOT-MWIR scope’s range is significantly farther than the current thermal weapon sights. Such a capability should lead to increased standoff distance for snipers and provide a significant advantage over adversaries.”

About the Author

Michael Hoffman
Michael Hoffman is the executive editor at Tandem NSI and a contributor to Military.com. He can be reached at mhoffman@tandemnsi.com.
  • dirtylodown

    The easier I can see you with US citizen.

  • GregSD

    How about this notion. We keep this kind of information out of the press and until we have placed it into action and tested the product we all shut the hell up. Our enemies will now make plans/change plans based on what we put in the media. We shut our mounths and maybe they won’t be ready for us and we can save some lives if we even have to confront them. What a novel concept. Unless this is something we are going to see in the commercial world, keep it quiet. DAMN!

    • Phono

      Eh, you have to see Propaganda as an even more powerfull “Weapon”. Just to think “they could possibly do that” will affect the enemy, and gives you an upper position psychologically.

      And … what will they do against a 10gpx-camera – what they don’t do against a 10mpx-camera?

    • Pilgrimman

      More like “Oh man, if they can do that and they are willing to de-classify it, God only knows what technology they have that is still classified!!”

      • blight_

        Not so sure of that. If the project is being put together at Duke, it is likely civilian funded at first before it trickles up to the military and back out to the civilian world.

    • footyfoot

      How about this notion- maybe this stuff isn’t nearly so secret and obscure as you might think it is? Artists have been experimenting with gigapixel formats for a few years now, as well as giant-format film which delivers incredible detail at long distances.

    • jeff

      Oh no, cameras. Hahaha you fucking idiot

    • Elias

      We learned this from the enemy in the first place , making it public is to let them know we have it … (its called phased array photography ) and darpa is within the trillion pixel already in the dungeon labs so its old news… (:P

  • davidz

    That tech would probably be usefull in heat seeking missiles too.

  • blight_

    They’re taking a page out of the use of distant arrays for astronomical imaging.

    Then again, it *is* time to slash NASA’s budget again…

    Edit: A concomitant advance in image recognition capability is needed to make this work, though having much higher quality images will help greatly. No use trying to make a mountain out of a molehill of five pixels. This isn’t the movies now…

  • Matt Holzmann

    The cameras in our spy satellites in the late 1960’s could read the license plates of Russian trucks from 150 miles up…

    • blight_

      I’ve always wondered if that were true, especially considering the technology of the ’60s.

      Do military vehicles even /have/ license plates? It sounds very civilian, and perhaps suspicious.

      • MarkN

        The advance isn’t the resolution; even assuming it was a meter at 250 miles up, the cameras used were neither lightweight nor low powered. In fact they were huge, hundreds of pounds huge, so huge as to be not adaptable to the battlefield. And what we are talking about here is something small and light enough to be weapons mounted, meaning ounces, not pounds.

        • blight_

          Well yeah.

          Even now we have no idea how big these Keyholes and their replacements are.

          The NRO may give up some old sats to NASA, and their paper capabilities are fairly impressive. IIRC, the speculation is that they’re basically Hubble Space Telescopes pointed at the ground.

        • Stormcharger

          Basically this. The difference is apples and oranges, Keyhole satellites do not use digital resolution. Their resolution capability is determined by aperture diameter and focal length, not how many megapixels it can resolve. Which means that with a large enough camera, it can focus on a very small area 200 to 400 miles away. A digital camera is much different in that the picture is made up of pixels on a screen, more pixels mean better resolution at a given distance because small objects can then be rendered more completely.

          And to note as well, it only takes about 20 minutes for a satellite to cross from horizon to horizon, which means that almost 50% of that time it is at an angle to see a vehicles licence plate.

      • EW3

        A license plate from 150 miles up ?

        Just consider the angles involved.
        From 150 miles straight up the license plate is not even visible.
        You need to see the plate from an angle, which means more atmosphere and possibly more trees, hills and mountains in the way.
        Very suspicious.

        • blight_

          Looking at Terrascale and other sat data, it’s rare to see angles that aren’t more than a few degrees from normal to the earth’s surface. You wouldn’t see drivers licenses to begin with.

          A U-2 is another story, not necessarily better.

          • mi1400

            seeing earth surface area and size of satellite/car, prbability of satellite being directly above it to fail reading … could be what 1% !?! …. but one thing is,… the distance added to 150miles because of angular view will require i guess 250miles capable camera!?!

      • chch

        In the ’70’s the US Dept of Interior had sats that could see a specific bug on a specific tree in a forest.

    • Thomas L. Nielsen

      Source, please.

      Regards & all,

      Thomas L. Nielsen
      Luxembourg

      • blight_

        Friend of a cousin of a friend?

  • Rob

    It’s tech like this we need tin America to compete with Asian imports in general. I want a chance for a 10 giagpixel cam

    Anyways, put these super cams on everything. Drones, carriers, outposts, minesweepers.

    Surveillance and ability to learn from scenarios recorded are a huge factor in today’s warfront.

    Hope this is not just hype.

  • Vaporhead

    Police State.

  • Mark

    Film was used before digital cameras. The resolution of the film is at the molecule size, far higher than any digital camera will ever be on a per shot basis. Here is a web page explaining this fairly well. I think military film size was 70 mm? Anyways here is the site. http://kenrockwell.com/tech/film-resolution.htm

    • Todd

      Film is exposed on a molecular level, yes. But the limits of film exposure rest in the optics. Distorting light to focus onto a piece of film is difficult. The edges of items being filmed are not imaged at a molecular level – rather there is a transition space which is larger or smaller depending on the optics used.

      The same effect is there with a digital system, since they rely on optics to focus the image on the digital image processor. While optics have been being developed for centuries, digital sensors have only been been in development for a few decades. It is, therefore, the combination of the two that determine what your image is capable of conveying for digital photography.

  • shawn1999

    My math may be off, but if a 10gpx camera/scope can read a plate at 540 yards, a 50gpx camera/scope can read at 2700 yards. Current record for longest range kill is at 2475 yards (according to Wikipedia- http://en.wikipedia.org/wiki/Longest_recorded_sni…. This record was achieved with a combination of great skill, excellent weaponry and perfect weather conditions (mountains of Astan) and is extremely difficult to reproduce at lower climes (Military Channel tried to recreate the 2nd longest- 2430 yds- shot also accomplished in Astan mountains, while in the AZ desert and could not spot the round to make adjustments- assuming this means the round was too far off trajectory and wouldn’t succeed anyway). This means the limitations of weapon/weather put the stand-off distance closer in. How does an enhanced camera increase the stand-off distance?

    • tmb2

      Snipers aren’t only used for killing – they’re used as scouts as well. The LRAS on the Cav Bradley can identify people at 10km, but the Bradley doesn’t possess a weapon that can shoot that far.

  • Lee

    We will develop it here…then we’ll send it to the Chicom’s to be manufactured, all the time convincing ourselves that they won’t steal our technology.

    This country is getting dumber and dumber…the Ruskies and Chicoms have got to be laughing their heads off.

  • Lothar Loc’nar

    Waiting on the smart weapons that recognize personnel from that far away and only hit them, singling them out with bio readings, it’s one thing to see a human and another to verify it’s the exact target.

    Also teleportation should definitely be on our agenda just tag the enemy and teleport him/her into the interrogation room.Tagged with nanobots and bio status/recognition.

    Wonder whatever happened to rail gun tech? (shrug)

    Also heat dispersion uniforms so the heat signature ammo cannot focus on our soldiers. With drones ability to p/u sweat emanations we will need a de-orderizer as well…lol

  • Matt Holzmann

    several of the comments here are spot on The enemy knew our capabilities because we would use photos to warn them that certain activities were known and that we would discourage such activities with certain escalations of scale or force.

    Zeiss Jena was kept open by the Russians because they were one of the very few places they could grind the lenses necessary for their own satellites, which were quite inferior but serviceable. Don’t know where the Chinese get their optics. This is all open source.

    and always assume the enemy is reading your mail.

  • JohnB

    The fridge size cam that weights 1/2 a ton. Chill down. Many DARPA projects ended wrapped in dirts and spider webs.

  • John Moore

    How about writing an article that shows a bit of technical literacy? 2000 times more powerful (repeated a few times) says nothing, since “power” is not defined. It’s hard to tell what the article is about. Is it about the high temp thermal imager, or a multi-gigapixel optical imager, or what?

    Sigh

  • William C.

    You know, it’s one thing to be vigilant of the government overstepping their boundaries, it’s another thing to shout “police state” at a breakthrough in camera/optic development. The first is reasonable, the second is crazy.

  • traindodger

    What our troops really need is a rifle scope with GPS/Gyro/Rangefinder abilities that can act as a target locator and designator. No more radioing in coordinates. If you can point at it with your rifle and lase it, you can call in CAS or indirect fire on it, uploading the coordinates to the battlefield network wirelessly and in real time.

    Combined with augmented reality, each soldier and vehicle crew member connected to the network would be able to see the locations being marked in their heads-up display, and react in kind with suppressive or direct-fire options.

    Oh, and since every scope would have a camera mounted in it, troops would be able to point their guns around corners and get a picture-in-picture view in their eyepiece of what their gun sees without exposing their body.

    Now THAT would be a game-changer. The trick is to make the electronics cheap enough.

    • tmb2

      Sounds like a lot of extra gear to weigh down a soldier. Give all that to a sniper’s spotter or a JTAC.

  • justajoe

    “A recent test found a camera could accurately read signs and license plates up to 270 yards away….The next step is to build a 10-gigapixel camera capable of reading license plates 540 yards away.”

    Really not very impressive distances… this story is probably missing some just info–just line many of the stories on here.

    • marcello

      the point is that without considering what kind of lenses are used with the camera it doesn’t make much sense to speak in terms of “reading license plates” up to some distance.

      with the right set of glasses even a 640×480 pixel webcam could be used to read license plates at quite some distance.

      • blight_

        Optics are the bottleneck for a CCD.

        I imagine that DARPA is paying for that as well…

  • wpnexp

    LOL, weren’t we reading license plates from space years ago, says the squint.

  • Brian

    The point is you could give this to a soldier and increase situational awareness. Is there a guy hiding in that building over there? Zoom in and get a better look.

  • blight_

    http://www.darpa.mil/NewsEvents/Releases/2012/07/

    “DARPA successfully tested cameras with 1.4 and 0.96 gigapixel resolution at the Naval Research Lab in Washington, DC. The gigapixel cameras combine 100-150 small cameras with a spherical objective lens. Local aberration correction and focus in the small cameras enable extremely high resolution shots with smaller system volume and less distortion than traditional wide field lens systems. The DARPA effort hopes to produce resolution up to 10 and 50 gigapixels—much higher resolution than the human eye can see. Analogous to a parallel-processor supercomputer, the AWARE camera design uses parallel multi-scale micro cameras to form a wide field panoramic image.”

    I think we’re not emphasizing the panoramic part of it as much as we should. Sure, one can develop high quality optics that can zoom, but at a severe situational awareness penalty.

    Combined with machine-learning systems to recognize distant objects, you could use it for target recognition or base security. And instead of employing clouds of cameras, all highly zoomed in but with terrible situational awareness, you have one camera that delivers high zoom panoramic images.

    Put it on a mast and use it for cav scouts. You acquire high quality imagery over a whole area, then zoom in to what you need, versus pushing up the mast and trying to find needles in a haystack.

    Another interesting possibility might be IED detection. A powerful panoramic camera could watch a swath of road ahead of a convoy, and image it to incredible fidelity. With panoramic mode, it wouldn’t just be imaging a piece of roadside but with wide field you would get the road and off-road, so while checking the road for characteristics of someone putting a IED you might be able to see the wire leading off towards the house nearby.

    And of course, with one camera you can image a huge swath of busy street and won’t be as vulnerable to blind spots from smaller, high-zoom, small-field-of-view surveillance cameras…

  • longshadow

    What’s the point of fielding a scope that outperforms its weapon? The ROI is a little sketchy too. This seems more applicable to small espionage drones than sniping platforms.