Wide Dynamic Range is generally considered a valuable feature with a few notable exceptions
By Colin Bodbyl
Wide dynamic range (WDR), is a feature that can significantly impact the image quality of a camera.
WDR is not new; in fact, it’s a term invented by the video surveillance industry to describe cameras with a high dynamic range.
There are several different methods for increasing the dynamic range of a camera but to understand how they work, and what is considered true WDR, you first need to understand what dynamic range is.
The dynamic range of a camera indicates how well the device can display both very dark and very bright areas of a scene at the same time. Most people have experienced the effects of a poor dynamic range while using personal cameras to take photos. The best example can be seen when taking a photo of someone from inside a building with a bright window behind them. The image will typically come out with the person visible but the area outside the window
completely white or otherwise washed out.
This is because the light levels inside the building are so dramatically different to the light levels outside the window that the camera can only balance to one or the other.
In personal imaging or photography cases, this limit in range is often desirable. It can create an appearance of realism and adds depth to the image, but when used in video surveillance it can be a problem. Using the same example
above, there is a risk the camera instead balances on the light level outside of the window, in turn, eliminating identifiable features of the person who would then appear like a silhouette in front of the window. This is where WDR allows surveillance cameras to capture both the background details outside the window as well as the identifying features of the person in front of it.
There are many methods for increasing the dynamic range of an image, but only one is broadly accepted as true WDR. This method involves the camera capturing multiple images at the same time or milliseconds apart and then combining them together to create a single WDR image.
For example, if a WDR camera was to capture an image of the person in front of the window, it would take three images. The first would balance on the background outside the window, the second would balance on the person in the foreground, and the third would capture an image somewhere between the two. The camera would then combine the most detailed areas of all three images to create a single WDR image. WDR is generally a valuable feature on almost any camera, except when using machine vision AI or video analytics. One benefit of a low dynamic range is that it creates higher contrast images with clear definition between objects which makes it easy for a computer to distinguish between them. WDR flattens the image by reducing the highlights and shadows, in turn creating an image with very little contrast (areas of pure white or black) that is difficult for machine vision AI to analyze. For this reason, it is
usually advisable when using video analytics to disable the WDR feature in the camera.
Outside of its impact on AI, one of the biggest challenges with WDR is that the feature has no standards or regulation around it, leaving manufacturers to self-declare whether a camera supports the feature or not. Since there are other
ways to increase the appearance of a cameras dynamic range, this can be very confusing for the buyers.
For the right applications, WDR can be a valuable tool for capturing as much detail as possible, but WDR has created a lot of confusion for buyers. This is compounded by its negative impact on some machine vision AI along with some manufacturers falsely claiming their cameras support the feature. Perhaps someday WDR will be regulated and certified. Otherwise, for now, the only way to determine a camera’s WDR capabilities is also the most effective way: test it in real life.