Business & Marketing
Ask the Expert: What’s the best way to ensure image quality?
By Robert Moore
Or have you ever thought you were getting a day/night camera only to find that it merely boosts gain and switches the colour image to black and white in order to “see” in the dark? With the lack of standards governing many aspects of video quality, purchasing a camera based solely on its datasheet is risky business. Instead, the best way to distinguish image quality between different cameras is by testing them side by side in challenging lighting conditions to compare three factors: visibility, appearance and fidelity.
By Robert Moore
A number of outside factors impact visibility, including direct sunlight, darkness, haze and smoke. If you’ve ever been stuck in traffic because of “solar glare” or gone out fishing early before the fog has lifted, you know that the human eye can’t discern much detail in these conditions. The good news is that video cameras can see better than humans in extreme conditions like these because of the technology. In fact, installers have several choices to improve visibility, including incorporating true day/night cameras, leveraging infrared (IR) illuminators, or even using thermal cameras.
True day/night cameras have the ability to see in the spectrum of light — namely infrared light — where we can’t. To be considered a true day/night camera, it must have an automatic IR cut filter that remains in front of the camera’s image sensor during daylight to filter out the infrared light that can distort an image, and is then retracted from the sensor to maximize the entire light spectrum when the camera detects low light conditions. Such cameras can display colour images with as little as .001 lux of light. Day/night cameras are ideal where artificial light is either not practical or desired at night, such as in shopping centres near residential areas.
You can achieve even better visibility in very dark conditions with the combination of IR illuminators and true day/night cameras. IR illuminators provide luminosity in the 800 to 1,000 nanometer range, as compared to the 400 to 750 nanometer range that humans can see. Additionally, IR illuminators can be pointed in the direction that you need, which means you can obtain ideal camera view angles without artificial light. But, as with any technology, there are limitations. For example, non-covert IR illuminators create a red aura that intruders can discern with the use of infrared detection devices. As an alternative, you can install covert IR systems, but they are more expensive. In addition, illuminators add to the overall cost of a system and increase power consumption considerably.
As a third alternative, thermal network cameras, which operate up to 14,000 nanometers, can be utilized. They detect and compare heat signatures from different objects to the surrounding environment to produce an image, which makes them better at detecting potential threats regardless of visibility issues. The fact that thermal technology can easily discern the differences between humans, animals and inanimate objects, they are very attractive for intelligent video applications because of a reduction in false alarms.
This is a more subjective aspect of image quality, and very often mistaken for visibility. Focusing on appearance can result in images that are more pleasing to the human eye.
You can create this by high contrast, exaggerated colours and artificial sharpening. If you increase those parameters in an image, you get a more pleasant picture that people often judge as a correct representation of the scene even when actual reality is, in fact, duller.
If you’ve ever seen a re-mastered, colourized version of “The Wizard of Oz” or an old Three Stooges episode, the actors’ rosy red cheeks are a perfect example of an exaggerated image.
Adjusting the image will also give you an impression of seeing more details, even if you don’t see more at all — it is just a subjective impression.
The final factor determining image quality is fidelity: how close does the colour captured by the camera match what the eye actually sees?
While visibility and appearance are more important factors on the surface, identifying the actual colour can be just as important in security, such as when you need to identify the colour of a perpetrator’s jacket in a sea of people or tell the difference between a $5 red chip and a $500 dollar pink chip.
Using HDTV standard image sensors that guarantee colour fidelity is vital here. With HDTV image sensors and intelligent video applications, you can, for instance, successfully track a suspect’s red jacket as he runs through a crowd. Without an HDTV sensor, objects may have colour artifacts that aren’t true to the image, and image colours may not be consistent across different monitors.
So instead of selecting your cameras based on datasheet facts, compare the factors of visibility, appearance and fidelity in a real-life situation. Remember that your ultimate goal is to match the surveillance needs of the customer rather than satisfy a checkbox in an RFP. By field testing network cameras and benchmarking them against each other, you’ll avoid surprises when it comes time to deploy.
Robert Moore is Canadian Country Manager, Axis Communications. Robert.Moore@axis.com