While they weren’t on my original list of exhibitors that I planned to visit, ICRealtime got my attention and immediately became my most interesting target. I kind of stumbled across Brian Levy at their booth. He was patient with my excited questions and my marveling at not having heard of them before in spite of the many awards they have won at fairly prestigious shows. I will admit that I have a personal interest in the burgeoning field of 360 capture. I’m sure there will be an agreed upon term for this new tech but at the moment, it goes by a few names: 360, VR, Immersive, etc.

The basic idea is that you have camera technology that is capable of capturing the views from almost every possible angle from a single device at the same time. I say “almost” as the camera itself takes up some space and all of the current models manage to exclude the device from view so you actually have a tiny bit less than “all” in your capture. But the primary bullet point is the same- you can have a single camera capture an entire room which eliminates both the need for multiple cameras and Pan/Tilt mechanisms. I could dive into the effect that viewing these types of captures through VR goggles or even Google Cardboard has but that is not relevant to this story so I’ll save the evangelizing for another day. ICRealtime had multiple options but I was focused on two models: the IC720 and the Allie. Before I get into the details of both cameras, there were two points that Brian made that I think are worth repeating. While this may sound like marketing fluff to some, I think it carries more weight for the majority of American consumers. First, both of their 360 camera models are completely made (not just assembled) in the US. Second, all of their underlying technology is protected by patents that they own.

 

The IC720 is the commercial model. It’s currently around the $700 mark for dealer pricing and has features such as:

  • Two view options – You can view the capture by navigating around the field or have things presented in a single view that provides all angels in one shot. Their software adjusts the image so the single-view mode is actually usable.  
  • High resolution sensors – 2 12MP fish eye lenses (25 MegaPixels). The feed is stitched together in real time and adjusted to optimize the presentation.
  • Focus agnostic analytics – Analytics are unaffected by where the user is actually looking. You can add bounding boxes to any point in the entire scene and triggers will be appropriately responded to even if the event happens 180 degrees behind where the navigated view frame is focused. Going one step further, the solution includes a feature where the view focus is automatically changed to the trigger event location so you don’t miss something happening in real time.
  • Compression improvements – customized H264 and AAC codecs to minimize the data output and reduce the demand on storage and bandwidth.
  • ONVIF compliant – keeps you from getting stuck in a walled garden.
  • Snapdragon 800 processor – when you are stitching together two fisheye feeds, you will need all the cpu power you can get.

ic720 from ICRealtime

I could just list the spec sheet but I doubt that’s worth your time. The main point is that the IC720 does with one camera what normally takes several for the purposes of security and surveillance. I asked Brian about who was buying these cameras and he responded first by mentioning some sports stadiums and schools (think gym coverage) but the general point was that anyone with a large area to cover can immediately see something to gain from leveraging this new solution. The uses for this technology extend far beyond simply big, open spaces but it will be the people with those needs who are likely to be the first adopters for 360 capture in surveillance. Sample Video

 

The Allie camera from ICRealtime is their consumer equivalent of the IC720. While it retails for around $500, it’s got enough in the features and hardware to justify the price. While the unit features 2 8MP fisheye lenses (instead of 12MP), the rest of the bullet list above is true for the Allie. Sample Video In addition, it includes:

  • 2 way audio
  • IR night vision
  • Cloud recording (ICMycloud)
  • Mobile app access (Android/iOS)

allie 360 degree security camera from ICRealtime

While Allie was intended for home use, they have prepared it for mobile capture. According to Brian, all you need is a 9V external battery (with USB cable) and a phone with signal and mobile hotspot feature. In fact, even if you don’t have a great cell signal, the camera will stream the captures to the mobile phone so you don’t lose anything due to network access problems.

One thing you notice with 360 cameras, in general, is that the still photos look much better than the video. This turns out to be an artifact of how the video information is ‘stretched’ around the virtual orb that provides the immersive playback. This means that you will have a natural driver to continue increasing the capture resolution over time; in theory. I asked about any plans to increase the resolution of their sensors given that the market, in general, is always seeking more pixels. To that Brian responded by pointing out the natural, technical obstacles that govern the speed with which higher resolution options come to market. Namely, considerations such as storage and bandwidth. If you need an upstream connection of ~1MB to service a standard HD consumer camera (without choking the entire connection) then the content generated by a 25MP camera is going to produce significantly greater demands on the NVR storage and required bandwidth. That is going to be a major consideration for customers looking to purchase. So, trying to increase the resolution of the captures too quickly does not pay off. Having said that, they are always working to address the impact of increasing the the amount of data being captured and processed. For example, codec tweaks to H264 and a plan to move to a customized version of H265.

So what’s so great about a 360 capture? In short, it’s how you can ‘view’ it. While the new solutions allow for a very novel method of capture, the method of consumption is where you find the real value. The single shot view where everything captured is visible in a single box is nice and potentially useful but definitely NOT what makes the experience of viewing this kind of capture compelling. It’s when you actually navigate ‘around’ the image that the value becomes apparent. It’s much like the effect of watching film. You have a roll of single photos arranged so that when you move them past a light and lens fast enough, the human brain stitches them together to create the illusion of movement. In that same way, moving your view around a 360 capture allows your brain to stitch together the shifting focus so you can create a sensation of immersion not possible with traditional technology. Beyond that, there are additional advantages over traditional surveillance cameras. For example, a traditional camera needs physical hardware (ie- servo motors) to adjust where the camera is focusing. This means a need for more power and a larger camera body (to house the moving parts) as well as producing a slow and comparatively laggy change in view. With the two, fisheye cameras from ICRealtime, the effect was silky smooth scene change that moved as fast as I could navigate. The Pan and Tilt are virtual since all views are being captured at the same time. This is why the navigation can be both smooth and as responsive as the user.

Was everything about the new products fantastic? Of course not. This is pioneer tech and, as a result, has some rough edges that are produced as a natural byproduct of carving out new territory. On close inspection, you can detect some artifacts at the very edges of the lenses but that doesn’t distort what matters. One example is the slightly fuzzy edge where the camera itself is removed from the shot. I’m sure that with the opportunity to test the equipment myself, I would find more shortcomings but that is both expected and beside the point. The camera solutions that ICRealtime is working to provide represent a new branch in the tree of surveillance technology.

An one might expect for anyone trying to develop a new area in the field, their results didn’t exactly match up with their “Plan A”. According to Brian, getting their solutions to where they are today took 2 years longer than the originally proposed 2 year development plan. The more familiar you are with the realities of getting a new technology to market, the less surprised you will be with such a delay but it reflects the slog that stands in between a great, new idea and actually having a product that customers can buy.  
They are winning awards but get very little (IMO) press. I can only speculate on why this is but regardless, I will be betting on this type of solution for the future. While there are, technically speaking, some vendors starting to get into this type of technology, ICRealtime has managed a decent head start.