Dallmeier Panomera Panorama
by Security Electronics and Networks | @Articles Product Reviews | August 8, 2011, 7:00am AEST
WHEN I first saw Panomera I was more than a little confused – it’s a rectangular box with rows of lenses poking out of it, like the broadside of some 17th-century battleship. Getting a handle on exactly what Panomera is capable of doing just by reading the blurbs posted online didn’t help much either.
But when I sat down with Vlado Damjanovski at C.R. Kennedy’s Mascot showroom, it started to make sense. The key thing to bear in mind with Panomera is that it allows objects very deep in a scene to be viewed in the same high resolution as those objects in the foreground. Or it allows exceedingly wide scenes to be viewed – or serious width and depth. And it allows simultaneous full resolution recordings of these scenes at up to 30ips. That’s pretty much Panomera in a nutshell.
Of course flat words don’t really have the impact Panomera technology did when I saw that first demo. Panomera allows end users to view scenes a thousand metres wide or a thousand metres deep with full resolution everywhere. On screen, such capability is highly impactful. You can pan from one end of a row of commercial aircraft to the other, then dig into a scene at any point and pull out a recognisable face while recording everything in full resolution.
While this technology has an optical heart, it’s the power of the Panomera software that brings it all together into a workable solution. That’s because the system is built on optical layers that are combined together by Panomera software to form a composite single image that can be built using up to 24 of Dallmeier’s 2.1 megapixel cameras giving a total scene resolution of 51MP. That total resolution is what can be viewed at a given moment as well as what is recorded all the time.
We’ve all been impressed by 10 and 20MP cameras before and there’s no doubt such cameras offer awesome performance. But Panomera is a different animal. Its scenes are wider and deeper than anything I’ve ever seen before and are achieved through the software-based amalgamation of raw optical horsepower.
“Optically, the cameras viewing furthest away have narrower view lenses and you match those so you can zoom in as if they are close to you – that’s where the trick is – that’s how Dallmeier is combining these scenes”
According to Damjanovski, Panomera first appeared at the Essen Show in Germany last year but he says Dallmeier’s people kept it quiet as they did not want to say too much at that point in its development.
“The Panomera solution went public in a big way at the IFSEC earlier this year, by winning the major award” he says. “My first impressions were that it was a heavy metal block, with multiple HD cameras attached, perfectly positioned and optically aligned for continuous wide optical view. But in reality it’s more than that. It’s promoted as 51 MP because you can put up to 24 cameras (2.1MP per camera) in a single Dallmeier multichannel box. With Dallmeier’s efficient H264 encoding all 24 can be stored and processed in a single box.”
As Damjanovski explains, Panomera is a solution that is electro-mechanically set up by Dallmeier to cover a given scene using a combination of a number of Dallmeier HD cameras whose images streams once captured are seamlessly stitched together using the Panomera software. Although, more correct term would be “combined” instead of “stitched” since stitching assumes overlapping of images, combined is adding to the view in seamless continuation.
Interestingly, Dallmeier builds the solution itself to supplied site specifications to ensure there’s no errors built into it and Damjanovski reckons that part of the installation would be the most challenging to do and adjust, but it will be done by the factory.
“A particular Panomera solution can incorporate up to 24 standard Dallmeier HD cameras – each with different fields of view and depths of field so as to offer complete coverage of a large scene,” Damjanovski says. “These images are combined by software to give a huge single image that can be mined for detail. Due to the highly efficient video compression Dallmeier uses, there’s now enough computer power to process the combined multiple images into a single image stream,” he says.
“So – there’s a hardware component and a software component. The hardware component is the box in which the cameras are installed and the cameras themselves – how many cameras you install depends on what you want to achieve.”
“From a camera point of view it’s the existing Dallmeier HD camera – it’s nothing different. It has a good lens and is optically matched to the scene. The Panomera software then processes image streams on the fly and analyses for motion allowing for automatic following and digitally zooming on the area of interest, without losing vision from the total wide scene, allowing for other operators to follow other areas of interest of the same wide view.”
“Panomera is recorded on any of the Dallmeier’s multichannel boxes, such as the Leonardo or Amadeus series – up to 24 HD cameras per box – that means you can build up a mosaic of cameras.”
Because Panomera is such a new release, the images Damjanovski is showing me are decoded image streams first, as seen at the IFSEC show, rather than streams from an actual Panomera solution. But these recordings are just as they would be if we were viewing such a system in real time.
“This exhibition solution included 12 cameras but a system can have 15, 8, 24 depending on the number of layers of vision you need to cover a scene,” explains Damjanovski.
“These layers of vision are created by different cameras in order to get different depths of field and fields of view. It looks continuous but you can see the areas the scenes are joined by the software – it’s subtle but it’s there.”
And the images are awesome – they are massive – We are viewing fields of view 1000 metres wide and more then drilling in to get high resolution images. It’s a striking effect.
First we take a look at images from a stadium and see Panomera’s amazing ability to pan across a broad horizontal scene in perfect focus, yet be able to burrow into the scene and ferret out detail 10 or 15 rows back – or further – depending on the configuration of a given Panomera solution.
Next, we look at a carpark. For comparison, we start out looking at the scene through a single MP camera. With one camera you get a decent view but there’s a point in the scene – early mid-depth – where resolution falls apart and it’s impossible to recognise faces, license plates – there’s simply too much pixelation. Then we switch to the Panomera and the whole scene becomes clear at all points as we move about. We can see all the way into the scene with perfect resolution and get face recognition easily.
According to Damjanovski, when it comes to covering a scene optically you can have any shape you want, rectangular or horizontal or halfway in between.
“Optically, the cameras viewing furthest away have narrower view lenses and you match those so you can zoom in as if they are close to you – that’s where the trick is – that’s how Dallmeier is combining these scenes,” he explains.
“It simply depends what you want to see – is it horizontal like a line of airside gates at an airport – a 180-degree scene? Is it narrow but deep like a large carpark, a public square or a railway platform? In a local application you could see the entire Sydney Harbour Bridge on a single screen and with detection zones you could establish events based on motion.
“This technology simply leverages the fact that every camera can give you full detail at a particular depth of field if you have narrow view lens – it’s very simple really,” he says. “It does not make sense in very small environments, like little shops, but in larger areas, indoor or outdoor, it’s very effective. It’s all IP, PoE too, so there’s only one cable required for video transmission. Power and comms to one pole.”
Recording the full scene at all times is a big advantage of Panomera. It goes without saying that with a PTZ you only see a small part of the scene with clarity – this is the part where the PTZ camera is pointing and focused on, but you can’t see and record the rest of the scene if the camera is not pointing that way. With Panomera you see the whole scene in great detail and can pan tilt and zoom in the scene and record the entire scene simultaneously. That’s the power of this solution. And the scenes are simply titanic.
“Another strong feature is that the Panomera system can be accessed by multiple users simultaneously so different users can get access to different areas of the image stream, depending on what they need to see,” Damjanovski explains.
Owing to the Multicast capability several users can view the images from Panomera without requiring repeated transmission of the data via network. This reduces the necessary bandwidth significantly.
Another thing that’s nice about Panomera is that it’s intrinsically scalable. The number of sensors/cameras in a solution can be adjusted depending on the needs of the site and they can be upgraded later to meet changing needs.
And there’s that allows the passages of people or vehicles to be automatically reconstructed over long distances with no gaps in the recording or camera switching.
According to Damjanovski, Panomera is a system that’s easy to put into practise and not too expensive to install on a site. From my perspective it’s also one of those solutions that proves the empiricist maxim – with Panomera, seeing really is believing.