The Major Milestones of IP Video

Sept. 28, 2012
Where we've been; Where we're headed

You know you have been around too long when people start asking you to write “Milestones” articles. Yet it seems like only yesterday that Axis Communications Co-Founder Martin Gren returned to Sweden from a Tokyo business trip with his idea for a network-connected video camera. Sixteen years later, IP video still possesses the spirit of a new technology thanks to all the fast-paced innovation and perpetual excitement it brings to the security industry.

IP video has grown from a web attraction device, to a niche, not-yet-ready-for-primetime player in the security world, to an enabler of 100,000-plus camera surveillance systems. We have seen technological innovations both big and small that have brought us to where we are today — but since a network camera at its core is an intelligent IP-enabled computing device that also happens to “see” — you must look beyond the camera’s video performance to fully appreciate each milestone.

While image and video quality features of the IP camera are often cited first and foremost, business functionality and system size milestones are equally important. These three timelines make the IP video market what it is today.

Image and Video Quality Milestones

Compared to the world of analog CCTV, where the last true video innovation was the emergence of color in the 1990s, IP video technology has delivered a steady stream of video quality improvements.

While the first IP camera’s performance was extremely modest, today’s security practitioners receive the same video quality in the control room that they have on their flat-screen TVs at home.

  • 1996 – World’s first network camera launched by Axis at the N+I event (today called Interop) in Atlanta. By today’s standards, the ‘Neteye’ had underwhelming performance – especially for video surveillance – highlighted by one frame every 17 seconds in normal resolution, and one frame per second in CIF resolution, which is less than 0.1 megapixel.
  • 2000 – First full frame rate network video camera, but only in CIF.
  • 2003 –First full frame rate network camera in 4CIF with the ability to use progressive scan technology. This gave IP video its first official – albeit small – advantage over analog cameras.
  • 2004 –Megapixel hits the network video market for the first time. With analog CCTV forever stuck at 540 Horizontal TV Lines (HTVL), this was the first major image quality advantage for IP, however it came with limited frame rate and limited light sensitivity.
  • 2006 – We see the first multi-megapixel IP cameras beginning to ship. More pixels meant more of an advantage over analog, but it also meant larger data packets, which limited frame rate in real world applications. Light sensitivity improved a bit, but most megapixel cameras used M-JPEG compression.
  • 2008 –Taking a lead from the consumer entertainment world, H.264 compression is brought to IP surveillance for the first time.
  • 2009 – Debut of the world’s first HDTV-compliant IP camera. By following the Society of Motion Picture and Television Engineers (SMPTE) standard, not only could IP video deliver the same 720p/1080i resolution that we enjoy in our living rooms, but the security industry realized the benefits of full frame rate high definition video, a 16:9 aspect ratio for better coverage and guaranteed color fidelity. Finally, IP video quality is better than analog in ALL aspects, and with 30 fps in 1080p resolution, IP cameras deliver 600 times better performance as compared to the first Neteye camera.
  • 2010 – The addition of the P-iris lens with precise control software inside the camera to address challenging lighting makes high megapixel cameras much more useful. Megapixel frames rates also become reasonable with up to 12 fps at 5MP. HDTV also spreads from the original fixed box camera to all form factors, including PTZ dome.
  • 2011 – Years of IP video imaging expertise combined with the latest in-camera chipsets and lensing lead to the newest color-at-night Lightfinder technology. Color video is now possible down to 0.05 lux (FYI, a full moon on clear night delivers about 0.1 lux).
  • 2012 –Wide Dynamic Range with dynamic capture comes to IP video. The advanced image processing leverages next generation in-camera chipsets with Main Profile H.264 to deliver several images with different exposure times up to 120 dB, resulting in video with exceptional clarity and sharpness in the toughest lighting conditions.

Increased Business Functionality Milestones

Unlike the steady innovation the industry enjoyed with image quality, the development of video intelligence for business functionality took a rockier road. Initially, we saw a nice adoption curve with the introduction of motion detection and I/O ports, but a combination of post-9/11 hype, reactionary government funding programs and before-its-time expectations caused a few false starts with intelligent video.

With millions of IP cameras in the world today, there is a real business need for intelligent video. Fortunately, we appear to be back on track and have the three ingredients for effective analytics: exceptional image quality, ever growing processing power and advanced algorithms. It’s likely that analytics will be a serious market driver for years to come.

  • 1998 – The first input/output (I/O) ports appear on a network camera, including the RS-232 serial standard. By enabling end-customers to integrate door contacts, alarms and even Point of Sale (POS) data with video, the novel idea of using network video for business operations was created.

  • 2000 – IP video intelligence is born with the introduction of Video Motion Detection (VMD) at the edge.

  • 2005 – Four years after 9/11, government reactionary funding for improved security and surveillance lead to the emergence of many intelligent software companies from out of their basements and garages. Big intelligent video hype begins, with new players in the surveillance market promising facial recognition and bag-left-behind analytics well before the technology was ready.

  • 2006 – A few bits of good news emerge from the brief analytic funding utopia with the introduction of practical video intelligence applications like People Counting and Camera Tampering, which is an extremely powerful feature of many IP cameras, albeit rarely used.

  • 2007 – Intelligent video hype continues as companies hanging around from the ’05 surge continue their work to convince customers that they can find that bag. Their biggest issue, perhaps, was that the mission critical analytic applications they promised required a near 100 percent accuracy rate. Even one false alarm or missed positive could prove disastrous.

  • 2009 –The recession is in full swing and the lack of venture funding effectively kills the intelligent video hype. In spite of the overall analytics hype, the extremely functional Audio Detection analytic makes its debut.

  • 2009/2010 – Improving in-camera chipsets provide enough processing power for the creation of analytic platforms at the edge for the software development and implementation of in-camera analytics. IP video “apps” now have a future.

  • 2011 – By following Moore’s Law, which is a natural IT innovation that enables twice the processing power every 18 months, chips launched in 2011 have the capacity to run nearly 95 percent of analytics at the edge.

  • 2012 – With the current limitations of professional-grade lenses and the fact that image quality has reached a point where few users require more resolution than what’s available today, the race for megapixels is over. The race for business and operational intelligence begins. With the ability to download and run applications directly at the edge, the IP camera is now the iPhone of the surveillance world


TCO, Scalability and System Size Milestones

One of the core benefits of any IP-based technology is scalability. The ability to add either 100 or just one camera to an existing system has always been a key total cost of ownership (TCO) benefit when comparing IP video with analog CCTV. In the early days of IP video, this scalability benefit proved extremely valuable for large systems.

However, as IP video manufacturers went Elephant Hunting for the biggest projects, there was less of a focus on developing technology for smaller systems, specifically 16-cameras-and-less, where analog had a strong foothold. This small system market is the final frontier for IP video.

  • 1998 – The first video encoder (also known as a video server) launches, with the goal of digitizing analog cameras in large systems. Additionally, the very first 10-camera IP video install is completed at a check cashing facility in New York City.

  • 2000 – The 2000 Summer Olympics is the catalyst for the first 100-camera IP video system at the Sydney Airport, built with 100 single-channel encoders.

  • 2002 – The government eRate program combined with a rise in Ethernet infrastructure and fear of school violence leads the Education market to become one of the earliest adopters and proponents of IP video. Atlanta Public Schools was an original pioneer, citing the lower cost of network video as a driver.

  • 2003 – IP video systems continue to grow, with the first 1000-camera systems. Cisco Systems has one of the largest installations to date with an all-Axis encoder, 2500 camera system.

  • 2007 – While IP video has proven to scale effectively in larger and larger systems, a TCO analysis conducted by Lund University in Sweden concludes that the breaking point for total cost benefit of network video over analog cameras is down to 32 cameras thanks to increased IP infrastructure, improved storage efficiency and common-off-the-shelf devices.

  • 2008 – 10,000-camera systems become operational, such as the one installed throughout the Stockholm Public Transport system.

  • 2010 – An updated TCO study shows that network video now offers a lower cost vs. analog when more than 25 cameras and, in some systems, even down to 14 cameras. At the same time, a major scalabiliy milestone is reached with the very first 100,000 network video camera system installed with a major retailer.

  • 2011 – While the concept of cloud computing is almost as old as the computer itself, and the ability for a hosted video system had been around since the mid-2000s, national systems integrators and enterprise-class cloud storage vendors partner to roll out video surveillance-as-a-service programs. Hosted video now offers an intriguing surveillance model for small systems, especially for business owners of many different locations.

  • 2012 – The concept of edge storage takes a leap forward with the introduction of all-edge recording solutions. While SD-card storage had been available for several years for redundant and critical storage, the price performance of an all-edge system with integrated software that eliminates the need for a separate recorder makes IP video viable for small systems.

While the IP video industry continues to focus on and deliver large system projects, the technology has started to scale down into smaller systems over the past two years. The recent milestones from the third timeline means that the technology milestones from the first two timelines – HDTV, intelligent performance, color-at-night video, apps, etc. – can now be enjoyed by small system owners.

In fact, of all the camera systems in the world, it’s estimated that a high majority are less than 16 cameras. Of that less-than-16-camera market, nearly 95 percent still use analog. It might be hard to fathom, but with the projected growth in front of us, IP video is still in its infancy with plenty more milestones to come.