How to Improve Anran's Person Detection Accuracy
The person detection feature on Anran security cameras is a powerful tool designed to reduce the number of false alerts you receive. By using artificial intelligence (AI) to analyse movement, the camera can differentiate between people, animals, vehicles, and other motion. However, if this feature isn't working correctly, you might miss important events or be bothered by irrelevant notifications.
This guide provides practical troubleshooting steps to help you fix issues with your Anran camera's person detection and improve its accuracy.
Common Person Detection Problems
You may be experiencing one of two main issues:
- False Positives: The camera alerts you for 'person detected' when it's actually an animal, a moving tree, or a shadow.
- False Negatives: A person walks clearly in front of the camera, but it fails to send a person detection alert.
These problems are usually caused by the camera's positioning, its settings, or environmental factors.
Step-by-Step Guide to Fixing Person Detection
Follow these steps to fine-tune your camera's performance.
1. Optimise Camera Placement and Angle
The way your camera is positioned has the biggest impact on the AI's ability to work correctly.
- Mounting Height: Install the camera at a height of 7-10 feet (2-3 metres). Mounting it too low or too high can distort the shape of a person, making them harder for the AI to recognise.
- Downward Angle: The camera should be angled downwards, not pointing straight out at the horizon. This provides a clear view of the person's entire body as they approach.
- Avoid Obstructions: Ensure there are no tree branches, bushes, or other objects that partially block the camera's view of the detection zone.
- Lighting: While the camera has night vision, very poor lighting or deep shadows can make detection more difficult. Ensure the area is as evenly lit as possible. A streetlight or porch light can be very helpful.
2. Adjust Detection Sensitivity
The sensitivity setting determines how much movement is required to trigger an analysis.
- Open the Anran app and select the camera you want to adjust.
- Go to Settings > Alarm Setting or Detection Settings.
- Find the Humanoid Detection or Person Detection sensitivity slider.
- If you are getting false alerts: Lower the sensitivity level.
- If the camera is missing people: Increase the sensitivity level.
- Make small adjustments and then test the results before adjusting again.
3. Define a Clear Motion Zone
Use the app to draw a specific activity zone. This tells the camera to only analyse motion within that defined area.
- Draw a zone that covers the area you want to monitor, such as your walkway or front door.
- Exclude areas with constant, irrelevant motion. For example, if a busy road is in the background, draw the zone to exclude the road. This prevents the camera from wasting processing power on cars, which can sometimes be mistaken for people at a distance.
4. Update the Firmware and App
Anran regularly releases software updates to improve performance and fix bugs. The AI algorithms for person detection are constantly being refined.
- Update the App: Check the Google Play Store or Apple App Store to ensure you have the latest version of the Anran app.
- Update the Firmware: In the camera's settings within the app, look for a 'Device Information' or 'Firmware Update' section. If an update is available, install it.
5. Reduce Environmental 'Noise'
Complex scenes can confuse the AI.
- Moving Shadows: Long, moving shadows during sunrise or sunset can sometimes be misinterpreted. While not always avoidable, be aware of this when positioning your camera.
- Swaying Trees: If you're getting false alerts from a specific tree, try to prune the branches or adjust the motion zone to exclude it.
By systematically optimising your camera's position and settings, you can dramatically improve the accuracy of the person detection feature, ensuring you get the alerts that matter.