Some of the emerging trends in CCTV surveillance feature the same IP camera technologies popularised in consumer electronics and broadcast video today. These technologies, including HDR (high dynamic range) imaging and H.265 video compression schemes, were highlighted recently at IFSEC 2015, Europe’s leading fire and security expo.
Here we reveal some of the advancements in camera technology for professional surveillance, some of which are familiar from consumer and broadcast video, and others such as video analytics that are somewhat specific to security applications.
Yes, I know, “CCTV” in many ways is an outdated term, but it’s still commonly used to describe surveillance video for security applications.
HDR: High Dynamic Range
While HDR currently is all the rage in the consumer and broadcast video markets – heck, even the iPhone has it – the technology is only just catching on in CCTV, in part because of its very high bandwidth requirements.
HDR enables the capture of fine details in both the shadows and highlights of any given scene.
In surveillance systems, you can imagine how the technology would be useful in discerning details that might help nab a perpetrator.
I saw very few HDR cameras at IFSEC (one of them was a new consumer product), but I did see a demo from the Taiwanese firm Dynacolor.
The company claims:
Implemented with the line-by-line HDR technology, instead of traditional frame-by-frame HDR method, DynaColor Z2V6 Series performs better image combination, lifting up HDR function to a higher level. This distinguishing feature corrects the backlighting problems in high contrast environments. Moreover, with high dynamic range functions supported, DynaColor Z2V6 Series captures more evidentiary details in high contrast environments.
Who knows what trickery was going on with the image capture, but in a side-by-side comparison at IFSEC, the HDR image of a bright light against a colourful cat sculpture and a patterned backdrop was far superior to the non-HDR image.
The Dynacolor HDR products in the company’s Z-Series cameras include the 2M HDR and 4M HDR.
Ultra HD up to 7K
When it comes to high-resolution, surveillance is still catching up to entertainment. CCTV is still touting boring old 4K UHD, while the TV hype is all around 8K.
There was at least one 7K camera at IFSEC from Vancouver-based Avigilon, which claims to have showcased “the security industry’s first single-sensor 7K (30 megapixel) security camera” in April at the ISC West 2015 in Las Vegas.
Avigilon also is now shipping its 6K (24 MP) HD Pro camera.
Throughout IFSEC we saw robotic cameras for every application from industrial disasters to consumer-oriented time-lapse recording.
At the same time, we saw a wealth of multi-lens 360/720-degree cameras to capture every angle at all times. So why the need for robotic cameras, which inevitably miss some of the action, no matter how fast the motor and how effective the sensors?
“There’s no distortion,” says Marcus Yang, CEO of Amaryllo, which makes a variety of consumer-oriented IP cameras.
The Amaryllo cameras feature “auto-tracking” via heat sensors that follow human activity. So what if more than one intruder invades the premises?
“How many people are going to break into your house?” replied one of the booth reps.
At CES 2015, Amaryllo demonstrated the iCam Pro, a wildly successful crowd-funded project that features 360-degree robotics, night vision, an SD card slot, three motion sensors, an LED light, two-way audio and other goodies. The unit is expected to ship at the end of this month for $300 (£190) retail.
At IFSEC, the company introduced the iSensor HD Patio – a 180-degree robotic camera and intruder alarm for the outdoors. The $199 (£125) features 4 GB of embedded memory and free 15 GB Google Drive storage.
So that’s the low-end consumer-oriented side of the motorized CCTV business. On the industrial side, there’s no stopping the robotics trends.
The e-vigilante from EOS – like an industrial-strength Roomba with a camera on top – motors around the premises looking for unusual activity and scouring the nooks and crannies that static cameras can’t catch.
High Frame Rates
Two years ago, Lilin launched what appears to be the industry’s first 120 fps surveillance camera. You’d think others would follow, but apparently they have not.
Lilin’s Jason Hill says he hasn’t seen such frame rates from other vendors. Lilin developed the solution for a couple of speciality applications such as number-plate capture, but since then the demand has stretched to other verticals including banking.
At IFSEC, Lilin demonstrated how its 120 fps camera could capture the details – down to the serial number – of bills flipping through a money counter.
He notes that the most sophisticated money counters should be able to detect a stray denomination, but in the event that the money counters miss, Lilin’s time-stamped recordings can determine when a wrong bill went through, helping the bank to find any glitches in its system.
Jason confesses that, out of secrecy, he doesn’t know what else Lilin’s banking customers are doing with the cameras: “They’re just telling us, ‘We need to be able to see the notes.’”
Jason notes that it’s hard for CCTV manufacturers to differentiate themselves in a crowded commercial market, but Lilin stands out for its willingness to “customise products to suit the market,” just as it has done for the residential custom-installation channel through integration with popular home automation systems.
Video Analytics is one camera technology that started with the security surveillance business for the purpose of counting people, recognizing faces, identifying and tracking objects, even detecting moods.
Mazi failed at interpreting my gender and age, but it got my emotion pretty much right: 32.6% happy.
One of the governing standards for video analytics, OpenCV (computer vision), has been exploited to extend analytics to many new applications, including such things as flame and smoke detection, enabling advanced warnings before traditional sensors detect heat and smoke-related chemicals.
It wasn’t too long ago that the H.264 video compression scheme was good enough for consumer electronics, TV broadcasts and CCTV transmission. But along came 4K and 8K resolutions, as well as the bandwidth hogging HDR.
Now all three sectors are making way for the latest version H.265, or High Efficiency Video Coding (HEVC), which was first published in 2013 and then revised in 2015.
HEVC is said to double the data compression of H.264/MPEG-4 AVC, while maintaining the same video quality.
CCTV manufacturers all over IFSEC touted cameras with the new technology.
Analytics + Compression
CCTV manufacturers are layering their own compression algorithms on top of industry standards like H.265. The goal is to optimise storage and bandwidth by recording higher-resolution video only where certain activities are detected.
If nothing is happening except trees swaying or paint drying, data might be captured in low resolution. But if humans or cars are detected in any given frame, those frames might be recorded in higher resolution.
While many makers of CCTV manufacturers offer their own flavours of such optimisation technology, they all sound pretty much like this description of Vivotek’s Smart Stream:
Smart Stream is a sophisticated technology developed by VIVOTEK in response to the conundrum that, all other factors being equal, the higher the video resolution, the greater amount of data produced, and thus the more network bandwidth consumed when transmitted. This situation forces many users to encode video from their network cameras at lower video quality in order to conserve network bandwidth, thus sacrificing surveillance quality.
Smart Stream is a technology developed by VIVOTEK to eliminate the need to choose between high quality video and efficient network usage, achieving this objective by providing higher video quality only in those areas in the camera’s field of view that are of interest to the user. These areas can either be predefined by the user as regions of interest (ROI), or detected automatically as moving objects. Background areas of no interest for surveillance are presented at a lower baseline video quality, thus saving on network bandwidth, as well as on storage capacity if the video is archived. The result is surveillance video that preserves video quality where it matters, while dramatically reducing overall network bandwidth consumption.
And this from Axis, describing its ZipStream technology:
Scenes containing interesting details are recorded in full image quality and resolution while other areas are filtered out, to optimally use available bandwidth and storage. Important forensic details like faces, tattoos or license plates are isolated and preserved, while irrelevant areas such as white walls, lawns and vegetation are sacrificed by smoothing in order to achieve the better storage savings.