Q: How to adjust the Motion Detection Areas of my WQHD Camera for an optimal Result?
A: For our older camera models we often recommended using a large Motion Detection Area to minimize false alerts. For our new camera series we improved the motion detection to be more reliable by surpressing background "noise" - see Background Compensation. This allows for a much more precise placement of the available detection areas.
Rules of Thumb when placing the detection areas:
- Place each area directly in the path that persons, animals or cars will take when crossing through the image.
- The detection areas do not only react when a person or a car is moving in front of the camera. But also sudden light changes - moving shadows, car headlights, etc - can trigger an alert. Try to place detection areas so that not affected by sudden changes in light conditions.
- Adjust the sensitivity of the detection according to the percentage of pixels inside the area you expect to change when you want an alarm to be triggered:
1 % sensitivity: 99% of the pixels inside the detection area have to change to trigger an alarm.
50 % sensitivity: 50% of the pixels inside the detection area have to change to trigger an alarm.
75 % sensitivity: 25% of the pixels inside the detection area have to change to trigger an alarm.
99 % sensitivity: 1% of the pixels inside the detection area have to change to trigger an alarm.
- The larger the detection area the more absolute pixels have to change to meet this condition. For example, an area covering the entire camera image will only trigger when something huge moves in front of the camera - or something small comes close enough to cover a large portion of the screen:
In the following image we see a person walking in front of the camera at a distance of approximately 5m. The detected motion is highlighted in white. And even though the sensitivity is set to
90% and no time we see a change of greater than
10% of the pixels - so we never get an alarm:
But now that we know the size of the expected object as well as it's expected position we can adjust the detection area accordingly. I also lower the sensitivity to
75% so that only
25% of the pixel need to change to trigger an alarm:
The placement and shape of the detection area now guarantees that a large amount of pixel will detect a change when another person will walk by our camera:
And this time the alarm was triggered successfully: