By: Jan van Essen
Multi-megapixels in security cameras. That must deliver excellent images. Or not? In this blog, I explain what role pixels play and which other factors determine the quality of a camera image.
Many people think that the more megapixels on the chip in the camera, the more details and the better the image. That is only partially correct. Depending on the situation and the intended purpose, there are also other factors that play a role.
Since digital (security) cameras made their entrance, the number of megapixels on the image sensor has increased enormously. From one million to 10 to 12 million pixels. The size of the image sensor, however, has (virtually) not changed. That is not possible either, because there is no room for it in today’s compact security cameras. We have started placing more pixels on the same surface, making the pixels themselves smaller. And this is where the problem lies. It appears that the smaller pixels can convert less light into an electrical signal, making them less suitable for dark situations. Security cameras intended for identification must not only see many details but also require high light sensitivity in dark situations. Most security risks occur during nighttime hours. The number of megapixels on the camera chip therefore only partially determines the quality of camera images.
Lens quality and size
Image quality is also determined by the quality and size of the lens used. To properly project image information onto the chip, a lens is needed that brings light from a certain angle of view together on the chip. The lens must also let as much light through to the chip as possible. This is done by using pure glass (which also prevents distortion) and making the lens as large as possible. Telephoto lenses on professional photo cameras have a lens the size of a cake plate for good reason. Unfortunately, such a lens size is not realistic on a security camera. The limitation of the smaller lens is visible when you digitally enlarge the image: it becomes blurry rather than looking at a blocky image as with low resolution.
With more pixels, we can maintain more detail with larger angles of view, but the limitations of the lens will then play a major role. A fisheye camera, for example, has a lens that projects the full 360 degrees of the situation onto the chip. Ideal for identification in smaller spaces such as entrances, stairwells, and elevators. When this camera is used in larger spaces such as parking garages and halls, you can only distinguish between, for example, a human or a dog at greater distances. Identification is no longer possible at that great distance. Currently, 12-megapixel fisheye cameras are already available for purchase, but they have a small lens with thick glass. And that is a limitation. If you want to create identifiable camera images in such large spaces, you would be better off choosing a multi-sensor camera. These are multiple megapixel camera modules with smaller angles of view in one housing, which transmits the images from the different modules as one image.
How many megapixels do you actually need?
Some camera manufacturers actively participate in the multi-megapixel hype and offer 4K (8 megapixel) or even 8K cameras. They have a hefty price tag and the sensitivity in the dark often leaves much to be desired. How many megapixels do you really need? An 8-megapixel image already provides, on a normal screen, the maximum detail that we can see with our eyes. Of course, we can still zoom in digitally, but then you also run into the optical limitations of the lens. With a fixed camera, you have many details, but in most situations, there is no need at all to count the nose hairs of the person at the back of the store. Moreover, the camera cannot look around corners. Multiple cameras with, for example, 3 megapixels (which are more light-sensitive) provide better image coverage and give a greater chance that the (un)desired situation is captured. With a controllable camera, you already bring the image optically forward and multi-megapixel is of secondary importance. It is nice to digitally zoom into the overview image afterward, but due to the lens package, a multi-zoom camera provides less good images.
Finally: all those megapixels also create large files and streams that must be sent over the network and stored. This disadvantage can be somewhat offset by applying image compression methods such as H.265.
Conclusion
When choosing suitable cameras, look especially beyond the number of (multi-) megapixels. Let your choice be determined primarily by the intended purpose. Multi-megapixel cameras are useful in large overview situations such as football stadiums, squares, and fields. But in virtually all situations, proper camera projection is finding a balance between need (including purpose and lighting conditions) and price, and multi-megapixels are certainly not the ultimate solution.