Features Business & Marketing
Top five considerations for choosing a network camera

Resolution

One of the key reasons an end-user chooses network video is because of improved image quality.  A network camera with SVGA resolution (800×600) has 40 per cent more pixels than a 4CIF camera, and a network camera with 5MP performance has 925 per cent, i.e. almost 10 times, more pixels than SVGA. If cost was a non-factor, you could simply choose the best resolution available. But, aside from the fact that higher resolution cameras have a price premium, there are two other “costs” to consider: First is the added cost of extra network traffic, and the second is the cost of the storage. The good news is that the costs for all are dropping while performance improves because of Moore’s Law as described in earlier articles.

In addition to analyzing the upfront and performance costs, you’ll need to determine what you want to see.  Is it just a general overview of an area? Will you need to zoom in for verification? Or will you need post forensic analysis of the image to determine finite details such as personal identification or a license plate number? Remember that your usable video data is linear — meaning either horizontal or vertical.  For instance, as you move from a 2MP to a 5MP camera, you won’t get a 1.5 times increase in usable video, but instead a 60 per cent increase in linear data. This means you can see the same object at a distance 60 per cent further instead of 150 per cent as you might intuitively think.

Also remember the value of HDTV standardized video, which follows the same SMPTE standard for video quality that television networks and movie studios follow. While the pixel resolution might be less than that of a higher megapixel camera, the overall video quality with respect to frame rate, aspect ratio, and color fidelity will create a more pleasing image to the eye.

Lighting

Lighting is a 24-hour problem, so you’ll need to know your environment before deciding between a standard network camera, a true day/night network camera, and a thermal network camera. 

For instance, if an indoor camera will have consistent office light all day, you can utilize a standard camera. However, if the area is going to have lower light conditions at a time when you still require visibility, you will want to have a camera with day/night capabilities.  A true day/night camera will have an automatic IR cut filter that blocks distortion-causing IR light during bright conditions, but is removed when it senses low light to allow all possible light in the visible spectrum to reach the camera’s sensor.  This will improve the camera’s image in the lower light conditions.

If you are installing in an area that will have a wide range of lighting – such as in a lobby that has sunlight pouring in through the windows during the day, but darkness at night – you will want to utilize a camera with Wide Dynamic Range. For outdoor installation, you will want to have both day/night capabilities and WDR to handle the ever-changing light.

Thermal network cameras are another type of camera that have become a popular option that work in conjunction with regular network cameras in both indoor and outdoor applications.  Thermal cameras are utilized in bright daylight so that you can get more accurate video analytics to detect activity as they are not fooled by shadows, tricky backdrops, or sunlight hitting the camera. They read heat signatures and have no video noise.  They are also perfect for pitch black situations.  Although applying artificial light is typically the best option, sometimes light pollution near a residential area or electricity costs are reasons to avoid adding light.  Remember, however, that thermal cameras detect activity but cannot identify the person of interest by the image – this is why they work best as a component of an overall surveillance solution.

Until the next Nikola Tesla comes along to perfect wireless power, this will continue to be a top consideration in camera selection.

Probably one of the most compelling reasons to choose a network camera is the ability to power the camera via Power-over-Ethernet (PoE).  In the early days of network video, you could only power fixed internal cameras because the IEEE standard 802.3af permitted 15 watts of power.  Pan/Tilt/Zoom cameras and cameras installed in cold weather climates like we have in Canada had to be powered traditionally with DC.

This has changed in the past year or two with the adoption of 802.3at that provides 25 watts over twisted pair Ethernet (UTP). There are also proprietary PoE solutions that allows for 50 watts and beyond.  So if you can reach the camera with Ethernet cabling, there is no reason not to use PoE for all of your network camera functions.

If you don’t have Ethernet capability, your choices are between using regular old power lines and utilizing solar energy. You will have to weigh the costs between trenching vs. installing costly solar.  In remote areas, solar may be your only choice.

Connectivity

A common question I get is, “Can I go wireless with your camera?”  The answer is yes, but I recommend that if you can reach the installation point with traditional wires, then go the traditional route.  Your two main choices for hard wiring the camera is Cat5e, Cat6 wiring or fibre optics.  If you have a camera that is less than 400 meters from the last leg in the connection you can go Cat5e/Cat6. Twisted pair Ethernet can go 400 meters by utilizing 3 PoE LAN extenders at 100 meter intervals.  This refutes the 100 meter myth often cited as a reason not to use IP video.

If you need to go even further, fibre optic is your choice.  One consideration is to go with fibre to the far location and have a switch at the end.  Assuming you have power for the switch you can connect multiple cameras at the far site with just one cable run essentially.

Still, if wired connections aren’t feasible, then any camera can “go wireless.” There are many choices to connect wirelessly through a camera’s Ethernet port.  It would take an entire article to explain the choices for wireless, but a quick list of the choices include 802.3a/b/g/n, point-to-point, non line of sight at 900Mhz, 5Ghz, etc., and wireless mesh.

Environment

Let’s talk temperature first. In Canada we are mostly concerned with the lowest temperature extremes, but you need to consider the higher temperature extremes as well.  A common temperature required for the northern latitudes is -40°C/F cameras.  It is interesting to note that even in Canada these temperatures are rarely reached – especially in the big cities.  A camera rated to -40° will survive even colder temperatures generally because of the heat generated by its internal electronics. There are manufacturers that create enclosures to keep cameras toasty to -70° (hello Antarctica!).

On the other extreme are cameras running in hot temperatures. A camera mounted on the roof in Phoenix will be much hotter than Phoenix’s 120°F temperature in the middle of a cloudless day.  This camera could melt if it doesn’t have protection.  A sunshield will help and in the hottest conditions you will want to add a fan as well as some air conditioning to keep the camera going.

IP-ratings are terms you may also hear in addition to temperature, and the most common for typical installations is IP66. This means that camera will resist dust and water. This is good for outdoors in Seattle, or indoors at a steel plant. But if you are, for instance, at a car wash with high pressure reaching the camera, you will need a higher IP rating. The first digit indicates the level of protection against solid foreign objects (i.e. dust) and the second digit indicates protection against water.

As you can see, some of these considerations will be the same from when you selected analogue cameras, but with IP technology, you have more options to choose from to create the best possible installation. By leveraging an open architecture, you can work with different partners to provide the products and education you need.