The latest camera of Axis Communications, the Axis Q1602, features improved technology designed to handle one of the most challenging problems of IP camera surveillance: capturing usable footage in low-light scenarios.
Axis say that the system will provide colour footage in light levels much lower than previous cameras, making people and objects easier to identify.
In this evaluation we are taking a look at the Q1602 to find out how well the camera copes where most cameras would struggle.
We looked at the footage provided by the Q1602 in low-light conditions to establish the benefits and drawbacks that the Lightfinder technology brings with it.
We did not test the camera’s capabilities with additional infrared or white-light lighting, and no outdoor testing was performed.
The problems with low-light surveillance
All conventional cameras require light to compose an image. The less light there is, the duller the picture until eventually nothing is shown. Modern IP cameras have a number of features which improve low-light sensitivity, the most obvious being infrared-sensitivity.
Most mid-to-high end cameras will feature a switchable infrared-cut filter for day and low-light monitoring. The benefit is that at night the camera is sensitive to infrared radiation. The drawback is that images will be monochrome, making it more difficult to identify people.
Another way of dealing with low light is lengthening the exposure time, increasing the light for each frame. While this does increase image clarity, any movement in view is smeared, making identification impossible for moving people or objects.
The Lightfinder technology isn’t designed to see in complete darkness, some light on the scene is still required. Axis themselves say “The LightFinder doesn’t create light, it only finds it”, so in areas of total darkness, this camera will need additional illumination.
First, we tested the camera at daytime light levels.
The image shows that in daylight, the images are on par with all conventional cameras.
As the light levels dropped, we took shots at regular intervals.
At light levels where the average camera would find it hard to capture a usable image, the Q1602 was coming into its own, with people and objects clearly defined and easy to identify.
One drawback we noticed, looking on the right-hand side of image 3, strong light sources are not compensated enough to avoid bleaching out areas within the image. In this case, the light is being produced by a monitor and the user is unidentifiable.
Then, we tested in near dark conditions.
At this very low light level, the camera is more sensitive than the human eye. In fact, we found it difficult moving around the office without bumping into desks. The video produced is still crisp and clear, with objects easy to identify.
Comparison with other cameras
We compared the Q1602 with another Axis camera, the P1346, without Lightfinder technology to see what the difference is.
We found that the P1346 camera automatically switched from day to night mode around 10lux, meaning only monochrome footage could be captured. The lack of colour is a hindrance on identification. Not being able to provide colour of key identifiers such as hair colour, clothing etc is a substantial disadvantage.
Finally, we compared the video captured by the Q1602 against the P1346 with someone moving at walking pace.
These shots highlight the massive benefit that the new Lightfinder feature provides in very low light. The person on the Q1602 is sharp and clear, while the P1346 is heavily blurred, to the extent that the footage is useless.
The reason for this effect is down to the automatically lengthened exposure time in the P1346. Between the start and the end of the exposure, the person has moved by around 20cm. As a result, the person is unidentifiable. In theory the exposure time could be limited to prevent this, but the reduction in light being captured would result in a black image. At either setting, the P1346 would not produce a useful image.
Importantly, with the Axis Q1602 the person clearly stands out from the background. This sharp contrast will improve the effectiveness and reliability of video analytics, such as video motion detection. With the dark colours and the blurring effect on the P1346, video motion detection is far more unreliable.
So what’s the drawback? The only drawback we found with the Q1602 is its resolution, which is limited to D1 (768×576). With most cameras sold these days featuring HD 720 / 1 megapixel resolution or higher, D1 resolution feels like a step back.
The law of physics rules: when sensor resolution increases, its’ sensitivity to light decreases since the same amount of light being captured is spread over more pixels. Keeping the resolution lower improves light sensitivity at the expense of detail. The Axis Q1602 is a prime example.
Axis’ claims about its Lightfinder technology are well founded. The difference in video quality and the usability of the footage produced in low light are substantial, making the Q1602 stand out from other security cameras when it comes to low light performance. The reproduction of colour in very low light conditions is not only impressive, it is very useful and can be critically important. Axis has set a new benchmark.
In addition to the benefits of improved colour and sharpness, there is a third, consequential benefit. Lighting is wasted when nothing is happening. And although the Axis Q1602 doesn’t eliminate the need for lighting, it significantly reduces the amount and power of lights required, saving both on the cost of equipment as well as cost and use of electricity.