Capturing High Performance
April 1, 2008
In the worlds of professional sports and entertainment, there are no clear definitions of what constitutes high performance. A high-performing baseball player may be a great pitcher or a slugging outfielder. A high-performing recording artist might be one who garners great reviews, packs concert halls or sells boatloads of CDs.
What’s true in sports and entertainment is no less true in the world of cameras designed for security applications. What determines “high performance”? According to experts contacted for this article, it can mean high resolution, low-light capability, rugged housing, wide-dynamic range, video analytics or any number of other qualities.
In this look at high-performance cameras, SDM talks with industry experts about their definitions of the term, how high performance and high definition differ, and how security dealers and integrators can best determine the appropriate high-performance cameras to specify and install.
The first point that experts stress is that high performance is a highly subjective term. In other words, it’s a term meaning different things to different users, with much depending on the environment and application in which the camera will be used.
“There’s no real industry definition I’m aware of for high-performance cameras,” says Don Beeby, vice-president of sales and marketing for the imaging technology division of CBC America Corporation in Commack, N.Y. “High-performance cameras incorporate the latest camera technology for a given application. It could be a standard analog camera. We have analog cameras in our line that are our biggest sellers.”
What definitions of high performance there are also can evolve over time, as technology itself evolves. So says Gary Perlin, vice president for video products with Speco Technologies in Amityville, N.Y., a 45-year-old firm specializing in all types of video surveillance equipment for residential, industrial and commercial uses.
To cite an example, years ago, it was very common for color cameras to offer 380 lines of resolution, Perlin says. When a new camera came out with 420 lines of resolution, that camera would have been termed high performance. A few years later, cameras were introduced with 480 lines of resolution, which became the new definition of high performance. Today, some cameras boast 520 to 560 lines of resolution, which has become the modern benchmark associated with high performance.
Resolution, however, is but one of the measures of high performance. “Another very important criteria is low-light sensitivity,” Perlin explains. “Over the years, low-light sensitivity has improved. At the same time all this is happening, they didn’t happen together. The highest resolution cameras do not have the best low-light sensitivity, and vice versa. So now you’re faced with the question of what you want the camera to do.
“The best camera for one application might be high resolution, while the best camera for another might feature low-light sensitivity. You have two cameras, both considered high performance, but each better for one application than another.”
Dan Scroggins, product marketing manager for Clovis, Calif.-based Pelco, says high-definition or megapixel cameras produce amazingly high-quality images in normal light situations, but struggle in low light. In security applications, Pelco often finds users have back-alley locations where exceptionally low-light capability is required. In that type of application, the best solution is a CCD imager with a lens boasting a very, very low F-stop, Scroggins says.
Pelco offers cameras producing visible images in 0.00018 lux, a yardstick used to measure light in a low-light scene. But in a no-light – rather than low-light – situation, these would not be considered high-performance cameras. “You would need a thermal imaging camera that actually detects IR radiation emitted as photons from molecules,” Scroggins says. “Thermal cameras we sell employ microbolometers, which collect those photons, process them and translate them into an image you can see on screen.”
Doug Marman, chief technology officer and vice president of products for Bedford, Mass.-based VideoIQ, also emphasizes the subjectivity of high performance.
“Sometimes it’s extreme environments; sometimes it’s illumination, and whether you can deal with no-light or low-light environments,” he says.
“A thermal camera would be considered a high-performance camera, because it can focus on body heat. And video analytics is another aspect, one that gives you performance you won’t find in a standard camera.”
HIGH PERFORMANCE VERSUS HIGH DEFINITIONHigh-performance cameras and high-definition cameras are not the same thing. High definition is often used in terms of home television and digital signals. However, CCTV is still 95 percent analog, and high definition and analog do not go hand-in-hand.
“But that’s changing,” Perlin notes. “And in five years, we’ll be more into a networked infrastructure where high definition becomes more important and doable.”
According to Beeby, analog cameras typically offer 540 lines of resolution, while high definition or megapixel cameras generally boast at least four times the resolution, and can go as high as 20 times the resolution. “So when you’re talking about high-definition cameras, you’re talking about megapixel cameras, in my mind,” he says.
THE CHARACTERISTICS OF HIGH PERFORMANCEIn addition to high resolution and low-light sensitivity, Perlin says a number of other characteristics of cameras may make them particularly appropriate for different settings and circumstances, and therefore should be considered high-performance cameras.
One is wide-dynamic range, which gives the camera the ability to see in mixed light scenes, where very light areas coexist with very dark areas. Cameras with wide-dynamic range can balance out these very light and very dark areas, and provide pictures that are pleasing when both kinds of light are present, Perlin explains.
Another characteristic some security professionals may seek is tamper-proof housing for the camera.
Still another trait is environmental stability, which is the capability to operate in temperature extremes and corrosive environments, such as an environment next to the ocean where salt water could corrode the camera and impair its functioning. And yet another way a camera might be considered high performance is if it has a varifocal or motorized zoom lens, as opposed to a standard fixed focal lens, according to Perlin.
For his part, Marman cites as an example of a high-performance camera one of his own company’s cameras with video analytics, a term referring to the camera’s automatic recognition of some aspect of the scene it is capturing.
Many users bemoan the fact their cameras don’t offer better resolution. But the cameras may have better resolution than what the users are recording. “By putting intelligence in the camera, we can allow them to record without taking up a lot of bandwidth or storage space, because they’re only recording something important,” he says. “This places this camera right in the middle, between high-performance cameras people are using without great resolution, and megapixel cameras that take up too much bandwidth and storage space.”
From his perspective as vice president of marketing and business development with Vancouver-based Avigilon, Pierre Parkinson believes in a perfect world, cameras would include both high definition and the ability to capture high dynamic range.
“The story always begins in the camera, and once an image leaves the camera, most of the surveillance systems apply lossy compression to the image,” Parkinson says. “And with that type of compression, the dynamic range and resolution gets degraded, so when you look at playback video, it’s so much worse than live video.”
Avigilon, he says, has designed an end-to-end high-definition system with what it is calling “high-definition stream management,” which preserves resolution and the dynamic range of the image.
WHAT TO LOOK FORThe features integrators should seek when selecting and installing a high-performance camera have much to do with the setting in which it will be used. Any good installer will begin by visiting the setting and conducting a site survey, in which he will evaluate the lighting conditions, environmental factors and requirements of the user, as well as the proper lens for the job, Perlin says. Based on that survey, he will specify the correct camera for that application and environment. “Consider that sometimes a simple, inexpensive camera is the right camera for the job, which in my mind makes that the high-performance camera for that application,” he adds.
In site surveys, integrators will want to measure the available light, the amount of available power, and how the videos will be transmitted, Scroggins says. They would also consider extreme conditions. “Deploy a camera on the north slope of Alaska, and you’ll need an extreme-environment-rated camera system,” he observes.
Beeby agrees with Perlin that not every job requires a high-performance camera offering state-of-the-art technology. Integrators look for a combination of performance and reliability at a given price when specifying cameras for any installation, he says.
Features they seek must go beyond the usual requirements like good resolution, ability to reproduce accurate colors and to handle difficult lighting situations. They need to also incorporate reliability, easy installation and a good range of mounting options, he reports.
A Megapixel PrimerAccording to Doug Marman of VideoIQ, several features distinguish megapixel cameras. “You start with an imager within the camera that can capture far more pixels of resolution than a standard camera,” he says. “Then you go into the processing, and there are differing approaches to compressing the video from the megapixel camera.”
Many megapixel cameras utilize jpeg compression, while some of the most recent megapixel cameras use MPEG-4 (referring to Motion Picture Experts Group, a different body than the one that established jpeg) for compression, Marman says.
There’s a tradeoff involved in going from jpeg to MPEG-4, he adds. The big step forward served up by MPEG-4 is that it uses a fraction of the bandwidth and storage space to provide the same quality of video. On the other hand, jpeg needs considerably less processing within the camera to achieve the compression.
And only recently have chips become available that allow MPEG-4 to compress videos from megapixel resolution, Marman explains.
The biggest advantage of megapixel cameras is that they offer more thorough coverage from the same camera. That, in turn, provides two distinct benefits.
“One is to allow you to essentially have one camera take the place of many cameras,” Marman says. “Let’s say you want to have enough resolution to identify license plate numbers passing along a multi-lane highway. A megapixel camera could give you the capability to read plates across several lanes. By contrast, a standard camera could only look at one lane with enough resolution to identify a license plate.”
The second benefit is the opportunity to use the camera to pan, tilt and zoom. A standard camera is likely to be able to see only so far in a field of view. But a megapixel camera can zoom in and see substantially farther.
Noting he believes the technology behind megapixel “is very misunderstood,” Speco Technologies’ Gary Perlin reports that megapixel cameras can only exist in network environments. It’s impossible to have an analog megapixel camera, he says. The most basic megapixel camera offers three times more resolution than an analog camera, and more advanced megapixel cameras can offer 10 times the resolution, he notes.
“The problem with megapixel cameras is there is so much information coming out of that camera that it takes a lot of time and bandwidth to transmit it,” he says. “That means you get fewer pictures per second. The upside of the megapixel camera is that there is so much detail that you can pan, tilt and zoom throughout the scene. That means that a fixed camera could be outfitted with an ultra-wide angle lens. But you can digitally manipulate the picture to get closeups within that ultra-wide view.”
So is it the camera or the lens that makes a megapixel camera a megapixel camera? According to Perlin, the answer is it’s the image sensor, or megapixel CCD (charge-coupled device). In addition, a number of components of the camera are enhanced to match the resolution the camera is capable of capturing. For instance, the DSP is upgraded, and a megapixel lens is used. “Putting a megapixel lens on a normal camera would not make it a megapixel camera,” he reports.
As Marman’s comments suggest, license plate identification is one of the chief applications for today’s megapixel technology. The other big application is enhanced facial recognition, according to CBC America Corporation’s Don Beeby who points to the March 2008 murder of a University of North Carolina student body president by an unknown assailant. Photos were taken at an automatic teller machine and a convenience store where the individual believed to be the student’s killer attempted to use her credit and ATM cards.
“In neither of these instances was the photo resolution clear enough to identify his face,” Beeby says. “In this case, a megapixel camera would likely have identified this individual.”
As for the cost premium involved in purchasing a megapixel, rather than a standard-resolution, networked camera, Perlin estimates the premium at about 30 percent. Moreover, the expense grows with the number of megapixels used, he says.
Marman reports that he has seen good quality megapixel cameras selling for about the same price as standard IP cameras. However, several of today’s megapixel cameras sell for two to three times that.
“One of the things to consider in purchasing a megapixel camera is the added bandwidth and storage costs,” Marman says. “You will not only have additional costs for the camera itself, but additional costs for the bandwidth and storage.”
Beeby’s estimate is that a megapixel camera will cost approximately four or five times as much as an analog camera, and two to three times as much as a standard resolution IP camera. “But prices are going to come down dramatically in the next year,” he adds. “It will pretty much follow the consumer market. All the consumer market camcorders are switching to megapixel now. And our industry typically lags the consumer market by a couple of years.”
When examining cost, remember that one megapixel camera may do the job of many standard analog cameras. For instance, Avigilon’s Pierre Parkinson reports that at the top of its camera line, its 16-megapixel camera offers the equivalent resolution of 52 standard analog cameras.