Keeping Up With High Definition
No matter what you call it — pixels, grids, horizontal lines or vertical mass — it still boils down to resolution. How much do you have? How much do you need? Can more be added? Will there be more in the near future? Will there be enough for everyone?
In the end, resolution is nothing more than a number. It is a digit we apply to an image to explain the amount of detail within it. Ask yourself a simple question: How defined is the picture I am looking at? Believe it or not, it's probably about the same amount today as it was 15 years ago. However, the definition of the picture is probably much better than it was five years ago. In the past few years, we have finally begun producing CCD cameras, recorders and monitors with resolution as good as their tube counterparts of 15 years ago.
The problem is we have pretty much reached our peak under the current NTSC and PAL standards. These are the two most popular broadcast standards in the world, and they dictate just how much bandwidth an image can use. Bandwidth relates directly to resolution — the more of one you have, the more of the other there is.
It's been a fairly well-known guideline: If you want to see the toys and tools of the future of CCTV industry, you should look at the current consumer industry. Recently, what was nearly a five-year time gap from adaptation of technology in one market to the other has dropped to less than two. So, using the current consumer market as a barometer, it seems the NTSC and PAL broadcast formats are about to be replaced by high-density television (HDTV) after NTSC and PAL's 50-plus years of dedicated service.
Get ready, because we are all just a few short steps away from brand-new televisions in our homes. At the very least, we could be required to attach converter boxes to our ancient televisions that will knock the resolution of the broadcast HDTV image back down to NTSC (525 vertical lines) or PAL (625 vertical lines) levels.
Here we are in the world of CCTV. We're excited about 400-plus horizontal line playback (0.5 million pixels). Meanwhile, the consumer industry is producing 4.2 million pixel resolution images (1000 +/- horizontal line res) from the same CCD formats. A 4.2 million pixel image is huge.
It's HDTV. It's the holy mackerel, all rolled up in a neat package.
But it isn't coming to the CCTV market for a while. Soon, my fellow fans, but not just yet. It may take three to seven more years (how's that for narrowing it down). Slowly, seriously improved resolution will become a seriously improved issue within our industry. Cost and compatibility with existing technology, however, remain important issues.
The bottom line is simple — camera technology is currently advancing, at a really quick clip. It's so quick, in fact, that the idea of a CCTV camera being able to replicate and/or exceed the capabilities of the human eye is no longer a dusty dream. But for now, resolution seems to be taking a back seat because of NTSC and PAL broadcast standards.
We need to be careful not to become too involved in keeping track of all of the new storage toys and innovations. We can't let our manufacturers step back 10 years and tell us that hugely improved resolution is not important to the industrial CCTV world. In other words, just because something isn't cost-effective today, we should not drop the ball for the near future. Where does that leave us? How about smack dab in the middle of a technology revolution?
Sure, digital storage systems are all the rage. But they are also being temporarily held at bay by analog recording systems, thanks to price drops, known stability, and, let's not forget, less-scary technical jargon to wade through during a simple purchase. This will continue for at least another year, maybe two. However, on the sidelines we are redefining the CCTV camera to become more than a glorified picture-taking machine.
Just a few years ago, we judged a camera based on its ability to provide a quality image in lower light. Of course, the lower light level, better-resolution images are still important, but they have become an almost everyday event. Today, we are categorizing our cameras into three different groups, and, for the most part, the groups have nothing to do with light levels.
Group one: IR sensitive devices
Step back and watch the failure of the past become the hero of the future. CMOS imagers are coming on strong, even though they were initially set aside because of their extremely high sensitivity to IR light, low sensitivity to poor lighting and high noise ratios. These chips were considered to be the cheap chip of the CCD industry, and they still are now. Found in almost every low-cost board camera, these units are being quickly adapted for higher uses because they are easily programmable and have high IR sensitivity. Keep in mind that we are not talking about IR cameras, just IR sensitivity. Yes, it's low-cost, low-sensitivity, low-resolution, high-noise programmable technology. Sounds like just the ticket when working close to your objective viewing target and/or the need for stealth using IR illuminators to gain low-light images. Don't worry, I've heard predictions that the CMOS imager will be going through some serious improvements over the next few years — watch out CCD.
Group two: Smart cameras
Here we go. Smart cameras are the digital revolution in the palm of your hand. Even though our cameras are being made smaller than my pinky finger, smart cameras are using on-board Digital Streaming Processors (DSP). The DSP allows us to make processing decisions based on the contents of the live video image at the camera! That means no special controllers, no wiring, no downloading — just good information processing. The DSP can also deliver better-quality images of areas with high contrast problems. That means more defined pictures of dark/bright areas and a wider dynamic range of light levels to work in. It also means we can build several different applications with extreme quality and response levels into the camera. Some of those applications can be:
· video compression using JPEG, MPEG, Wavelet, H263 or more; • on-site audio input compression;
· visual recognition applications (event and/or object driven);
· high-quality motion and/or non-motion detection;
· local image frame storage (continuous or event); and
· integrated communications for remote operation and image retrieval over a network.
Think about the savings that are staring us in the face! Instead of $10,000 worth of equipment, processors, controllers, and everything else, we do it all right at the camera. All we do is slap it up, program it, and let it go.
Group three: Network-compatible cameras
With the sudden promise and popularity of cameras that can be attached directly to a Local Area Network (LAN) versus running a whole bunch of extra cables, new network cameras are meeting the challenge. Want to go to a Wide Area Network (WAN), or perhaps the Internet using DSL or PSTN services? No problem. Plug-and-play surveillance is a reality. The main problem here is the need to always work with a local hard copy and pay attention to the size of the image and how many pictures-per-second are being sent and/or received. Like the smart cameras, these units come packed with good stuff: image and audio compression, local frame storage (both before and after event triggers), and motion and non-motion detection. Additionally, some units will include alarm inputs right on the camera, used for door contacts, photo beams and passive IR detectors, just to name a few.
There are more groups of cameras, and I have only opened the door just a bit. High resolution from a camera is no longer a problem. But high-resolution storage and high-resolution transmission are a different ball game. But we are, at least, on the field. Don't get into a big rush, because you should verify, test, prove and watch your designs and purchases carefully.
Keep in mind however, that we truly do live in the time of imagination — both within the mind of the purchaser and the seller. So, remember, it's buyer beware!
By Charlie R. Pierce
LeapFrog Training & Consulting