Abus Person Detection Wrong? How to Improve Accuracy
Abus security cameras with smart person detection are designed to provide you with meaningful alerts, filtering out irrelevant motion from things like passing cars or swaying trees. However, if you are receiving alerts when no one is there, or the camera is missing actual human movement, your settings may need some fine-tuning. This guide will help you troubleshoot and optimise your camera for accurate person detection.
The accuracy of this AI-driven feature depends on three key factors: camera placement, environmental conditions, and software configuration.
## Step 1: Optimise Camera Placement and Angle
Where you install your camera has the biggest impact on its ability to correctly identify people. A poor vantage point can easily confuse the detection algorithm.
- Installation Height and Angle: Install the camera between 2.5 to 3 metres (8-10 feet) high, angled slightly downwards. This provides a clear, natural view of a person's shape and size. Avoid placing the camera too high, which can distort proportions, or too low, where it can be easily triggered by small animals.
- Avoid Complex Backgrounds: Do not point the camera directly at areas with constant, non-human motion. This includes busy roads, flags, or trees and bushes that move a lot in the wind. This 'video noise' can sometimes be misinterpreted by the algorithm.
- Backlight and Lighting: Avoid pointing the camera directly at a strong light source, like the sun or a security light. Strong backlighting can create silhouettes, making it very difficult for the camera's AI to recognise a human form. Ensure the target area is as evenly lit as possible, especially at night.
## Step 2: Configure Detection Zones and Sensitivity
Use the camera's built-in settings to tell it exactly where to look and how sensitive it should be. This is crucial for eliminating false alarms.
- Log into Your Camera: Access your camera's settings through the corresponding Abus mobile app or web interface.
- Create Motion Detection Zones: Navigate to the 'Event' or 'Detection' settings. Look for an option to set 'Motion Detection Areas' or 'Activity Zones'. Draw zones specifically around the areas you care about, such as a doorway or pathway. Crucially, exclude areas with known triggers, like a busy pavement or your neighbour's garden.
- Adjust Sensitivity: In the same menu, you will find a 'Sensitivity' setting, often on a scale from 1 to 100. If you are getting too many false alerts, reduce the sensitivity level. A setting between 40-60 is often a good starting point.
- Enable Humanoid/Person Detection: Ensure the specific AI-powered filter for detecting people is enabled. This feature works in conjunction with general motion detection to analyse if the moving object has a human shape.
## Step 3: Rule Out Environmental and Other Factors
Sometimes, the cause of false alerts can be surprising.
- Insects and Spiders: At night, spiders are attracted to the warmth of the camera and the infrared (IR) lights. A spider crawling directly over the lens is a very common cause of false motion alerts. Regularly clean the camera lens and housing.
- Reflections and Shadows: On sunny or windy days, the moving shadows of trees or reflections from windows can be large enough to trigger a motion event. This is where carefully drawn detection zones become essential.
- Firmware Updates: Ensure your camera's firmware is up to date. Manufacturers constantly refine their detection algorithms, and installing the latest firmware can lead to a significant improvement in accuracy.
By systematically working through the placement, configuration, and environmental factors above, you can dramatically reduce incorrect person detection alerts and ensure your Abus camera notifies you only when it matters.